TruCluster Server
Hardware Configuration
Part Number: AA-RHGWB-TE
April 2000
Product Version:
TruCluster Server Version 5.0A
Operating System and Version: Tru64 UNIX Version 5.0A
This manual describes how to configure the hardware for a TruCluster
Server environment. TruCluster Server Version 5.0A runs on the Tru64™
UNIX® operating system.
Compaq Computer Corporation
Houston, Texas
Contents
About This Manual
1 Introduction
1.1
1.2
1.3
1.4
1.4.1
1.4.1.1
1.4.1.2
1.4.1.3
1.4.1.4
1.5
The TruCluster Server Product .....................................
1–1
1–2
1–3
1–3
1–3
1–3
1–4
1–4
1–5
1–5
Overview of the TruCluster Server Hardware Configuration ..
Memory Requirements ...............................................
Minimum Disk Requirements ......................................
Disks Needed for Installation ..................................
Tru64 UNIX Operating System Disk .....................
Clusterwide Disk(s) .........................................
Member Boot Disk ..........................................
Quorum Disk .................................................
Generic Two-Node Cluster ...........................................
Growing a Cluster from Minimum Storage to a NSPOF
Cluster ..................................................................
Two-Node Clusters Using an UltraSCSI BA356 Storage
Shelf and Minimum Disk Configurations ....................
Two-Node Clusters Using UltraSCSI BA356 Storage Units
with Increased Disk Configurations ..........................
Two-Node Configurations with UltraSCSI BA356 Storage
Units and Dual SCSI Buses ....................................
Using Hardware RAID to Mirror the Clusterwide Root
File System and Member System Boot Disks ................
Creating a NSPOF Cluster .....................................
Overview of Setting up the TruCluster Server Hardware
Configuration ..........................................................
1.6
1–7
1–8
1.6.1
1.6.2
1.6.3
1.6.4
1–10
1–12
1–13
1–15
1.6.5
1.7
1–17
2 Hardware Requirements and Restrictions
2.1
2.2
2.3
2.4
2.4.1
2.4.2
2.5
TruCluster Server Member System Requirements ..............
Memory Channel Restrictions ......................................
Fibre Channel Requirements and Restrictions ..................
SCSI Bus Adapter Restrictions .....................................
KZPSA-BB SCSI Adapter Restrictions .......................
KZPBA-CB SCSI Bus Adapter Restrictions .................
Disk Device Restrictions .............................................
2–1
2–1
2–3
2–6
2–6
2–6
2–7
Contents iii
2.6
2.7
2.8
2.9
2.10
RAID Array Controller Restrictions ...............................
SCSI Signal Converters ..............................................
DS-DWZZH-03 and DS-DWZZH-05 UltraSCSI Hubs ...........
SCSI Cables ............................................................
SCSI Terminators and Trilink Connectors ........................
2–7
2–8
2–9
2–9
2–11
3 Shared SCSI Bus Requirements and Configurations Using
UltraSCSI Hardware
3.1
3.2
Shared SCSI Bus Configuration Requirements ..................
SCSI Bus Performance ...............................................
SCSI Bus Versus SCSI Bus Segments ........................
Transmission Methods ..........................................
Data Path .........................................................
Bus Speed .........................................................
SCSI Bus Device Identification Numbers ........................
SCSI Bus Length ......................................................
Terminating the Shared SCSI Bus when Using UltraSCSI
Hubs ....................................................................
UltraSCSI Hubs .......................................................
Using a DWZZH UltraSCSI Hub in a Cluster
3–2
3–3
3–4
3–4
3–5
3–5
3–5
3–6
3.2.1
3.2.2
3.2.3
3.2.4
3.3
3.4
3.5
3–7
3–8
3.6
3.6.1
Configuration .....................................................
DS-DWZZH-03 Description ................................
DS-DWZZH-05 Description ................................
3–9
3–9
3.6.1.1
3.6.1.2
3–10
3–10
3–12
3–13
3–15
3–15
3–15
3–16
3.6.1.2.1
3.6.1.2.2
3.6.1.2.3
3.6.1.2.4
3.6.1.2.5
3.6.1.3
DS-DWZZH-05 Configuration Guidelines ..........
DS-DWZZH-05 Fair Arbitration .....................
DS-DWZZH-05 Address Configurations ............
SCSI Bus Termination Power .........................
DS-DWZZH-05 Indicators .............................
Installing the DS-DWZZH-05 UltraSCSI Hub ..........
Preparing the UltraSCSI Storage Configuration ................
Configuring Radially Connected TruCluster Server
Clusters with UltraSCSI Hardware ...........................
Preparing an HSZ70 or HSZ80 for a Shared SCSI Bus
Using Transparent Failover Mode ........................
Preparing a Dual-Redundant HSZ70 or HSZ80 for a
Shared SCSI Bus Using Multiple-Bus Failover ........
3.7
3.7.1
3–17
3–18
3–22
3.7.1.1
3.7.1.2
4 TruCluster Server System Configuration Using UltraSCSI
Hardware
4.1
4.2
Planning Your TruCluster Server Hardware Configuration ...
Obtaining the Firmware Release Notes ...........................
4–2
4–4
iv Contents
4.3
4.3.1
TruCluster Server Hardware Installation ........................
Installation of a KZPBA-CB Using Internal Termination
for a Radial Configuration ......................................
Displaying KZPBA-CB Adapters with the show Console
Commands ........................................................
Displaying Console Environment Variables and Setting
the KZPBA-CB SCSI ID ........................................
Displaying KZPBA-CB pk* or isp* Console
4–5
4–7
4.3.2
4–10
4–14
4.3.3
4.3.3.1
Environment Variables .....................................
Setting the KZPBA-CB SCSI ID ..........................
KZPBA-CB Termination Resistors ........................
4–15
4–17
4–17
4.3.3.2
4.3.3.3
5 Setting Up the Memory Channel Cluster Interconnect
5.1
Setting the Memory Channel Adapter J umpers ................
MC1 and MC1.5 J umpers .......................................
MC2 J umpers .....................................................
Installing the Memory Channel Adapter ..........................
Installing the MC2 Optical Converter in the Member System
Installing the Memory Channel Hub ..............................
Installing the Memory Channel Cables ...........................
Installing the MC1 or MC1.5 Cables ..........................
Connecting MC1 or MC1.5 Link Cables in Virtual Hub
Mode ...........................................................
Connecting MC1 Link Cables in Standard Hub Mode .
Installing the MC2 Cables ......................................
Installing the MC2 Cables for Virtual Hub Mode
5–2
5–2
5–3
5–5
5–6
5–6
5–7
5–7
5.1.1
5.1.2
5.2
5.3
5.4
5.5
5.5.1
5.5.1.1
5–8
5–8
5–9
5.5.1.2
5.5.2
5.5.2.1
Without Optical Converters ...............................
Installing MC2 Cables in Virtual Hub Mode Using
5–9
5–10
5–10
5.5.2.2
5.5.2.3
5.5.2.4
5.6
Optical Converters ..........................................
Connecting MC2 Link Cables in Standard Hub Mode
(No Fiber Optics) ............................................
Connecting MC2 Cables in Standard Hub Mode Using
Optical Converters ..........................................
Running Memory Channel Diagnostics ...........................
5–10
5–11
6 Using Fibre Channel Storage
6.1
6.2
6.2.1
6.2.2
6.2.2.1
Procedure for Installation Using Fibre Channel Disks .........
Fibre Channel Overview .............................................
Basic Fibre Channel Terminology .............................
Fibre Channel Topologies .......................................
Point-to-Point ................................................
6–2
6–4
6–4
6–5
6–6
Contents v
6.2.2.2
6.2.2.3
6.3
Fabric .........................................................
Arbitrated Loop Topology ..................................
Example Fibre Channel Configurations Supported by
TruCluster Server .....................................................
Fibre Channel Cluster Configurations for Transparent
Failover Mode .....................................................
Fibre Channel Cluster Configurations for Multiple-Bus
Failover Mode .....................................................
Zoning and Cascaded Switches .....................................
Zoning ..............................................................
Cascaded Switches ...............................................
Installing and Configuring Fibre Channel Hardware ...........
Installing and Setting Up the Fibre Channel Switch ......
Installing the Switch ........................................
Managing the Fibre Channel Switches ..................
6–6
6–7
6–8
6.3.1
6.3.2
6–8
6–10
6–13
6–13
6–14
6–15
6–15
6–16
6–17
6–17
6.4
6.4.1
6.4.2
6.5
6.5.1
6.5.1.1
6.5.1.2
6.5.1.2.1
6.5.1.2.2
Using the Switch Front Panel ........................
Setting the Ethernet IP Address and Subnet Mask
from the Front Panel ...................................
6–18
6.5.1.2.3
Setting the DS-DSGGB-AA Ethernet IP Address
and Subnet Mask from a PC or Terminal ...........
Logging Into the Switch with a Telnet Connection
Setting the Switch Name via Telnet Session .......
6–20
6–20
6–21
6.5.1.2.4
6.5.1.2.5
6.5.2
Installing and Configuring the KGPSA PCI-to-Fibre
Channel Adapter Module .......................................
Installing the KGPSA PCI-to-Fibre Channel Adapter
Module ........................................................
Setting the KGPSA-BC or KGPSA-CA to Run on a
6–22
6.5.2.1
6.5.2.2
6–22
Fabric .........................................................
Obtaining the Worldwide Names of KGPSA Adapters
Setting up the HSG80 Array Controller for Tru64 UNIX
Installation ........................................................
Obtaining the Worldwide Names of HSG80 Controller
Preparing to Install Tru64 UNIX and TruCluster Server on
Fibre Channel Storage ...............................................
Configuring the HSG80 Storagesets ..........................
Setting the Device Unit Number ..............................
Setting the bootdef_dev Console Environment Variable ...
Install the Base Operating System ................................
Resetting the bootdef_dev Console Environment Variable ..
Determining /dev/disk/dskn to Use for a Cluster Installation .
Installing the TruCluster Server Software .......................
Changing the HSG80 from Transparent to Multiple-Bus
Failover Mode .........................................................
6–23
6–25
6.5.2.3
6.5.3
6–26
6–31
6.5.3.1
6.6
6–33
6–33
6–40
6–46
6–48
6–49
6–51
6–53
6.6.1
6.6.2
6.6.3
6.7
6.8
6.9
6.10
6.11
6–54
vi Contents
6.12
Using the emx Manager to Display Fibre Channel Adapter
Information ............................................................
Using the emxmgr Utility to Display Fibre Channel
6–59
6.12.1
6.12.2
Adapter Information .............................................
Using the emxmgr Utility Interactively ......................
6–59
6–61
7 Preparing ATM Adapters
7.1
7.2
7.3
7.4
ATM Overview ........................................................
Installing ATM Adapters ............................................
Verifying ATM Fiber Optic Cable Connectivity ..................
ATMworks Adapter LEDs ...........................................
7–1
7–3
7–4
7–6
8 Configuring a Shared SCSI Bus for Tape Drive Use
8.1
Preparing the TZ88 for Shared Bus Usage .......................
Setting the TZ88N-VA SCSI ID ...............................
Cabling the TZ88N-VA ..........................................
Setting the TZ88N-TA SCSI ID ...............................
Cabling the TZ88N-TA ..........................................
Preparing the TZ89 for Shared SCSI Usage ......................
Setting the DS-TZ89N-VW SCSI ID ..........................
Cabling the DS-TZ89N-VW Tape Drives .....................
Setting the DS-TZ89N-TA SCSI ID ...........................
Cabling the DS-TZ89N-TA Tape Drives ......................
Compaq 20/40 GB DLT Tape Drive ................................
Setting the Compaq 20/40 GB DLT Tape Drive SCSI ID ..
Cabling the Compaq 20/40 GB DLT Tape Drive .............
Preparing the TZ885 for Shared SCSI Usage ....................
Setting the TZ885 SCSI ID ....................................
Cabling the TZ885 Tape Drive ................................
Preparing the TZ887 for Shared SCSI Bus Usage ...............
Setting the TZ887 SCSI ID ....................................
Cabling the TZ887 Tape Drive ................................
Preparing the TL891 and TL892 DLT MiniLibraries for
Shared SCSI Usage ...................................................
Setting the TL891 or TL892 SCSI ID .........................
Cabling the TL891 or TL892 MiniLibraries .................
Preparing the TL890 DLT MiniLibrary Expansion Unit .......
TL890 DLT MiniLibrary Expansion Unit Hardware .......
Preparing the DLT MiniLibraries for Shared SCSI Bus
Usage ..............................................................
8–1
8–2
8–3
8–4
8–4
8–5
8–5
8–7
8–8
8–8
8–9
8–9
8–10
8–13
8–13
8–13
8–15
8–15
8–16
8.1.1
8.1.2
8.1.3
8.1.4
8.2
8.2.1
8.2.2
8.2.3
8.2.4
8.3
8.3.1
8.3.2
8.4
8.4.1
8.4.2
8.5
8.5.1
8.5.2
8.6
8–18
8–18
8–20
8–23
8–24
8.6.1
8.6.2
8.7
8.7.1
8.7.2
8–24
Contents vii
8.7.2.1
8.7.2.2
8.7.2.3
8.7.2.4
8.8
Cabling the DLT MiniLibraries ...........................
Configuring a Base Module as a Slave ...................
Powering Up the DLT MiniLibrary .......................
Setting the TL890/TL891/TL892 SCSI ID ...............
Preparing the TL894 DLT Automated Tape Library for Shared
SCSI Bus Usage .......................................................
TL894 Robotic Controller Required Firmware ..............
Setting TL894 Robotics Controller and Tape Drive SCSI
IDs ..................................................................
TL894 Tape Library Internal Cabling ........................
Connecting the TL894 Tape Library to the Shared SCSI
Bus .................................................................
Preparing the TL895 DLT Automated Tape Library for Shared
SCSI Bus Usage .......................................................
TL895 Robotic Controller Required Firmware ..............
Setting the TL895 Tape Library SCSI IDs ...................
TL895 Tape Library Internal Cabling ........................
Upgrading a TL895 ..............................................
Connecting the TL895 Tape Library to the Shared SCSI
Bus .................................................................
Preparing the TL893 and TL896 Automated Tape Libraries
for Shared SCSI Bus Usage .........................................
Communications with the Host Computer ...................
MUC Switch Functions .........................................
Setting the MUC SCSI ID ......................................
Tape Drive SCSI IDs ............................................
TL893 and TL896 Automated Tape Library Internal
Cabling .............................................................
Connecting the TL893 and TL896 Automated Tape
8–24
8–26
8–28
8–28
8–30
8–30
8.8.1
8.8.2
8–30
8–33
8.8.3
8.8.4
8–34
8.9
8–36
8–37
8–37
8–38
8–40
8.9.1
8.9.2
8.9.3
8.9.4
8.9.5
8–40
8.10
8–40
8–42
8–42
8–43
8–43
8.10.1
8.10.2
8.10.3
8.10.4
8.10.5
8–44
8.10.6
8.11
Libraries to the Shared SCSI Bus .............................
Preparing the TL881 and TL891 DLT MiniLibraries for
Shared Bus Usage ....................................................
TL881 and TL891 DLT MiniLibraries Overview ............
TL881 and TL891 DLT MiniLibrary Tabletop Model ..
TL881 and TL891 MiniLibrary Rackmount
8–47
8–48
8–48
8–49
8.11.1
8.11.1.1
8.11.1.2
Components .................................................
TL881 and TL891 Rackmount Scalability ...............
DLT MiniLibrary Part Numbers ..........................
Preparing a TL881 or TL891 MiniLibrary for Shared SCSI
Bus Use ............................................................
Preparing a Tabletop Model or Base Unit for
8–49
8–50
8–51
8.11.1.3
8.11.1.4
8.11.2
8–52
8.11.2.1
Standalone Shared SCSI Bus Usage .....................
8–52
viii Contents
8.11.2.1.1
Setting the Standalone MiniLibrary Tape Drive
SCSI ID ..................................................
Cabling the TL881 or TL891 DLT MiniLibrary ....
Preparing a TL881 or TL891 Rackmount MiniLibrary
for Shared SCSI Bus Usage ................................
Cabling the Rackmount TL881 or TL891 DLT
8–53
8–54
8.11.2.1.2
8.11.2.2
8–58
8.11.2.2.1
8.11.2.2.2
MiniLibrary .............................................
Configuring a Base Unit as a Slave to the
8–58
Expansion Unit .........................................
Powering Up the TL881/TL891 DLT MiniLibrary .
Setting the SCSI IDs for a Rackmount TL881 or
8–61
8–62
8.11.2.2.3
8.11.2.2.4
TL891 DLT MiniLibrary ...............................
Compaq ESL9326D Enterprise Library ...........................
General Overview ................................................
ESL9326D Enterprise Library Overview .....................
Preparing the ESL9326D Enterprise Library for Shared
SCSI Bus Usage ..................................................
ESL9326D Enterprise Library Robotic and Tape Drive
Required Firmware .........................................
Library Electronics and Tape Drive SCSI IDs ..........
ESL9326D Enterprise Library Internal Cabling .......
Connecting the ESL9326D Enterprise Library to the
Shared SCSI Bus ............................................
8–63
8–64
8–64
8–65
8.12
8.12.1
8.12.2
8.12.3
8–65
8.12.3.1
8–66
8–66
8–66
8.12.3.2
8.12.3.3
8.12.3.4
8–68
9 Configurations Using External Termination or Radial Connections
to Non-UltraSCSI Devices
9.1
Using SCSI Bus Signal Converters ................................
Types of SCSI Bus Signal Converters .........................
Using the SCSI Bus Signal Converters .......................
DWZZA and DWZZB Signal Converter Termination ..
DS-BA35X-DA Termination ...............................
Terminating the Shared SCSI Bus .................................
Overview of Disk Storage Shelves ..................................
BA350 Storage Shelf .............................................
BA356 Storage Shelf .............................................
Non-UltraSCSI BA356 Storage Shelf ....................
UltraSCSI BA356 Storage Shelf ..........................
Preparing the Storage for Configurations Using External
Termination ............................................................
9–2
9–2
9–3
9–3
9–4
9–5
9–8
9–9
9.1.1
9.1.2
9.1.2.1
9.1.2.2
9.2
9.3
9.3.1
9.3.2
9.3.2.1
9.3.2.2
9.4
9–10
9–10
9–13
9–14
Contents ix
9.4.1
Preparing BA350, BA356, and UltraSCSI BA356 Storage
Shelves for an Externally Terminated TruCluster Server
Configuration .....................................................
Preparing a BA350 Storage Shelf for Shared SCSI
Usage ..........................................................
Preparing a BA356 Storage Shelf for Shared SCSI
Usage ..........................................................
Preparing an UltraSCSI BA356 Storage Shelf for a
TruCluster Configuration ..................................
Connecting Storage Shelves Together ........................
Connecting a BA350 and a BA356 for Shared SCSI
Bus Usage ....................................................
Connecting Two BA356s for Shared SCSI Bus Usage .
Connecting Two UltraSCSI BA356s for Shared SCSI
Bus Usage ....................................................
Cabling a Non-UltraSCSI RAID Array Controller to an
Externally Terminated Shared SCSI Bus ....................
Cabling an HSZ40 or HSZ50 in a Cluster Using
9–15
9–15
9–16
9.4.1.1
9.4.1.2
9.4.1.3
9–17
9–17
9.4.2
9.4.2.1
9–18
9–20
9.4.2.2
9.4.2.3
9–21
9–24
9–25
9–27
9–28
9.4.3
9.4.3.1
9.4.3.2
9.4.4
External Termination .......................................
Cabling an HSZ20 in a Cluster using External
Termination ..................................................
Cabling an HSZ40 or HSZ50 RAID Array Controller in a
Radial Configuration with an UltraSCSI Hub ..............
10 Configuring Systems for External Termination or Radial
Connections to Non-UltraSCSI Devices
10.1
TruCluster Server Hardware Installation Using PCI SCSI
Adapters ................................................................
Radial Installation of a KZPSA-BB or KZPBA-CB Using
Internal Termination ............................................
Installing a KZPSA-BB or KZPBA-CB Using External
Termination .......................................................
Displaying KZPSA-BB and KZPBA-CB Adapters with the
show Console Commands .......................................
Displaying Console Environment Variables and Setting
the KZPSA-BB and KZPBA-CB SCSI ID .....................
Displaying KZPSA-BB and KZPBA-CB pk* or isp*
Console Environment Variables ...........................
Setting the KZPBA-CB SCSI ID ..........................
Setting KZPSA-BB SCSI Bus ID, Bus Speed, and
10–1
10–2
10–6
10–9
10–13
10.1.1
10.1.2
10.1.3
10.1.4
10.1.4.1
10–13
10–16
10.1.4.2
10.1.4.3
Termination Power ..........................................
KZPSA-BB and KZPBA-CB Termination Resistors ....
10–17
10–18
10.1.4.4
x
Contents
10.1.4.5
Updating the KZPSA-BB Adapter Firmware ...........
10–18
A Worldwide ID to Disk Name Conversion Table
Index
Examples
4–1
4–2
4–3
4–4
4–5
Displaying Configuration on an AlphaServer DS20 .............
Displaying Devices on an AlphaServer DS20 ....................
Displaying Configuration on an AlphaServer 8200 ..............
Displaying Devices on an AlphaServer 8200 .....................
Displaying the pk* Console Environment Variables on an
AlphaServer DS20 System ..........................................
Displaying Console Variables for a KZPBA-CB on an
4–10
4–12
4–13
4–13
4–15
4–6
AlphaServer 8x00 System ...........................................
Setting the KZPBA-CB SCSI Bus ID ..............................
Running the mc_cable Test ..........................................
Determine HSG80 Connection Names ............................
Setting up the Mirrorset .............................................
Using the wwidmgr quickset Command to Set Device Unit
Number .................................................................
Sample Fibre Channel Device Names .............................
Displaying Configuration on an AlphaServer 4100 ..............
Displaying Devices on an AlphaServer 4100 .....................
Displaying Configuration on an AlphaServer 8200 ..............
Displaying Devices on an AlphaServer 8200 .....................
Displaying the pk* Console Environment Variables on an
AlphaServer 4100 System ...........................................
Displaying Console Variables for a KZPBA-CB on an
4–16
4–17
5–13
6–29
6–34
4–7
5–1
6–1
6–2
6–3
6–43
6–45
10–9
10–10
10–11
10–12
6–4
10–1
10–2
10–3
10–4
10–5
10–13
10–6
AlphaServer 8x00 System ...........................................
Displaying Console Variables for a KZPSA-BB on an
10–15
10–7
AlphaServer 8x00 System ...........................................
Setting the KZPBA-CB SCSI Bus ID ..............................
Setting KZPSA-BB SCSI Bus ID and Speed ......................
10–15
10–16
10–17
10–8
10–9
Figures
1–1
Two-Node Cluster with Minimum Disk Configuration and No
Quorum Disk ..........................................................
1–6
Contents xi
1–2
1–3
1–4
1–5
1–6
Generic Two-Node Cluster with Minimum Disk Configuration
and Quorum Disk .....................................................
Minimum Two-Node Cluster with UltraSCSI BA356 Storage
Unit .....................................................................
Two-Node Cluster with Two UltraSCSI DS-BA356 Storage
Units ....................................................................
Two-Node Configurations with UltraSCSI BA356 Storage
Units and Dual SCSI Buses .........................................
Cluster Configuration with HSZ70 Controllers in Transparent
Failover Mode .........................................................
NSPOF Cluster using HSZ70s in Multiple-Bus Failover Mode
NSPOF Fibre Channel Cluster using HSG80s in Multiple-Bus
Failover Mode .........................................................
VHDCI Trilink Connector (H8861-AA) ............................
DS-DWZZH-03 Front View ..........................................
DS-DWZZH-05 Rear View ...........................................
DS-DWZZH-05 Front View ..........................................
Shared SCSI Bus with HSZ70 Configured for Transparent
Failover .................................................................
Shared SCSI Bus with HSZ80 Configured for Transparent
Failover .................................................................
TruCluster Server Configuration with HSZ70 in Multiple-Bus
Failover Mode .........................................................
TruCluster Server Configuration with HSZ80 in Multiple-Bus
Failover Mode .........................................................
KZPBA-CB Termination Resistors .................................
Connecting Memory Channel Adapters to Hubs .................
Point-to-Point Topology ..............................................
Fabric Topology ........................................................
Arbitrated Loop Topology ............................................
Fibre Channel Single Switch Transparent Failover
1–7
1–9
1–11
1–13
1–14
1–16
1–7
1–8
1–17
3–8
3–10
3–14
3–15
3–1
3–2
3–3
3–4
3–5
3–20
3–21
3–24
3–6
3–7
3–8
3–25
4–18
5–9
6–6
6–7
4–1
5–1
6–1
6–2
6–3
6–4
6–8
Configuration .........................................................
Multiple-Bus NSPOF Configuration Number 1 ..................
Multiple-Bus NSPOF Configuration Number 2 ..................
Multiple-Bus NSPOF Configuration Number 3 ..................
A Simple Zoned Configuration ......................................
Emulated LAN Over an ATM Network ............................
TZ88N-VA SCSI ID Switches .......................................
Shared SCSI Buses with SBB Tape Drives .......................
DS-TZ89N-VW SCSI ID Switches ..................................
Compaq 20/40 GB DLT Tape Drive Rear Panel ..................
Cabling a Shared SCSI Bus with a Compaq 20/40 GB DLT
Tape Drive .............................................................
6–9
6–11
6–12
6–13
6–14
7–3
8–2
8–4
8–6
8–10
6–5
6–6
6–7
6–8
7–1
8–1
8–2
8–3
8–4
8–5
8–12
xii Contents
8–6
8–7
8–8
8–9
Cabling a Shared SCSI Bus with a TZ885 ........................
TZ887 DLT MiniLibrary Rear Panel ...............................
Cabling a Shared SCSI Bus with a TZ887 ........................
TruCluster Server Cluster with a TL892 on Two Shared SCSI
Buses ....................................................................
TL890 and TL892 DLT MiniLibraries on Shared SCSI Buses .
TL894 Tape Library Four-Bus Configuration ....................
Shared SCSI Buses with TL894 in Two-Bus Mode ..............
TL895 Tape Library Internal Cabling .............................
TL893 Three-Bus Configuration ....................................
TL896 Six-Bus Configuration .......................................
Shared SCSI Buses with TL896 in Three-Bus Mode ............
TL891 Standalone Cluster Configuration .........................
TL881 DLT MiniLibrary Rackmount Configuration ............
ESL9326D Internal Cabling ........................................
Standalone SCSI Signal Converter ................................
SBB SCSI Signal Converter .........................................
DS-BA35X-DA Personality Module Switches ....................
BN21W-0B Y Cable ...................................................
HD68 Trilink Connector (H885-AA) ...............................
BA350 Internal SCSI Bus ...........................................
BA356 Internal SCSI Bus ...........................................
BA356 J umper and Terminator Module Identification Pins ...
BA350 and BA356 Cabled for Shared SCSI Bus Usage .........
Two BA356s Cabled for Shared SCSI Bus Usage ................
Two UltraSCSI BA356s Cabled for Shared SCSI Bus Usage ..
Externally Terminated Shared SCSI Bus with Mid-Bus HSZ50
RAID Array Controllers .............................................
Externally Terminated Shared SCSI Bus with HSZ50 RAID
Array Controllers at Bus End .......................................
TruCluster Server Cluster Using DS-DWZZH-03, SCSI
8–15
8–16
8–17
8–23
8–26
8–33
8–35
8–39
8–45
8–46
8–48
8–57
8–60
8–67
9–4
9–4
9–5
9–7
9–8
9–10
9–12
9–13
9–19
9–21
9–23
8–10
8–11
8–12
8–13
8–14
8–15
8–16
8–17
8–18
8–19
9–1
9–2
9–3
9–4
9–5
9–6
9–7
9–8
9–9
9–10
9–11
9–12
9–26
9–27
9–30
9–13
9–14
9–15
Adapter with Terminators Installed, and HSZ50 ................
TruCluster Server Cluster Using KZPSA-BB SCSI Adapters,
a DS-DWZZH-05 UltraSCSI Hub, and an HSZ50 RAID Array
Controller ..............................................................
KZPSA-BB Termination Resistors .................................
9–31
10–18
10–1
Tables
2–1
2–2
2–3
AlphaServer Systems Supported for Fibre Channel ............
RAID Controller SCSI IDs ..........................................
Supported SCSI Cables ..............................................
2–4
2–7
2–10
Contents xiii
2–4
3–1
3–2
3–3
3–4
Supported SCSI Terminators and Trilink Connectors ..........
SCSI Bus Speeds ......................................................
SCSI Bus Segment Length ..........................................
DS-DWZZH UltraSCSI Hub Maximum Configurations ........
Hardware Components Used in Configuration Shown in
Figure 3–5 Through Figure 3–8 ....................................
Planning Your Configuration .......................................
Configuring TruCluster Server Hardware ........................
Installing the KZPBA-CB for Radial Connection to a DWZZH
UltraSCSI Hub ........................................................
MC1 and MC1.5 J umper Configuration ...........................
MC2 J umper Configuration .........................................
MC2 Linecard J umper Configurations ............................
Telnet Session Default User Names for Fibre Channel
2–11
3–5
3–7
3–11
3–21
4–3
4–6
4–1
4–2
4–3
4–9
5–2
5–3
5–5
5–1
5–2
5–3
6–1
Switches ................................................................
Converting Storageset Unit Numbers to Disk Names ..........
ATMworks Adapter LEDs ...........................................
TZ88N-VA Switch Settings ..........................................
DS-TZ89N-VW Switch Settings ....................................
Hardware Components Used to Create the Configuration
Shown in Figure 8 — 5 ...............................................
TL894 Default SCSI ID Settings ...................................
TL895 Default SCSI ID Settings ...................................
MUC Switch Functions ..............................................
MUC SCSI ID Selection .............................................
TL893 Default SCSI IDs .............................................
TL896 Default SCSI IDs .............................................
TL881 and TL891 MiniLibrary Performance and Capacity
Comparison ............................................................
DLT MiniLibrary Part Numbers ...................................
Hardware Components Used to Create the Configuration
Shown in Figure 8–17 ................................................
Hardware Components Used to Create the Configuration
Shown in Figure 8–18 ................................................
Shared SCSI Bus Cable and Terminator Connections for the
ESL9326D Enterprise Library ......................................
Hardware Components Used for Configuration Shown in
Figure 8–9 and Figure 8–10 .........................................
Hardware Components Used for Configuration Shown in
Figure 9–11 ............................................................
Hardware Components Used for Configuration Shown in
Figure 8–12 and Figure 8–13 .......................................
6–21
6–40
7–6
8–3
8–6
6–2
7–1
8–1
8–2
8–3
8–12
8–30
8–37
8–43
8–43
8–44
8–44
8–4
8–5
8–6
8–7
8–8
8–9
8–10
8–51
8–51
8–11
8–12
8–57
8–61
8–68
9–20
9–24
9–27
8–13
8–14
9–1
9–2
9–3
xiv Contents
9–4
Hardware Components Used in Configuration Shown in
Figure 9–14 ............................................................
Configuring TruCluster Server Hardware for Use with a PCI
SCSI Adapter ..........................................................
Installing the KZPSA-BB or KZPBA-CB for Radial Connection
to a DWZZH UltraSCSI Hub ........................................
Installing a KZPSA-BB or KZPBA-CB for use with External
Termination ............................................................
Converting Storageset Unit Numbers to Disk Names ..........
9–30
10–2
10–4
10–1
10–2
10–3
A–1
10–7
A–1
Contents xv
About This Manual
This manual describes how to set up and maintain the hardware
configuration for a TruCluster Server cluster.
Audience
This manual is for system administrators who will set up and configure the
hardware before installing the TruCluster Server software. The manual
assumes that you are familiar with the tools and methods needed to
maintain your hardware, operating system, and network.
Organization
This manual contains ten chapters and an index. The organization of this
manual has been restructured to provide a more streamlined manual.
Those chapters containing information on SCSI bus requirements and
configuration, and configuring hardware have been split up into two sets of
two chapters each. One set covers the UltraSCSI hardware and is geared
towards radial configurations. The other set covers configurations using
either external termination or radial connection to non-UltraSCSI devices.
A brief description of the contents follows:
Chapter 1
Introduces the TruCluster Server product and provides an overview
of setting up TruCluster Server hardware.
Chapter 2
Chapter 3
Describes hardware requirements and restrictions.
Contains information about setting up a shared SCSI bus, SCSI
bus requirements, and how to connect storage to a shared SCSI
bus using the latest UltraSCSI products (DS-DWZZH UltraSCSI
hubs, HSZ70 and HSZ80 RAID array controllers).
Chapter 4
Describes how to prepare systems for a TruCluster Server
configuration, and how to connect host bus adapters to shared
storage using the DS-DWZZH UltraSCSI hubs and the newest
RAID array controllers (HSZ70 and HSZ80).
Chapter 5
Chapter 6
Describes how to set up the Memory Channel cluster interconnect.
Provides an overview of Fibre Channel and describes how
to set up Fibre Channel hardware.
Chapter 7
Provides information on the use of, and installation of, Asynchronous
Transfer Mode (ATM) hardware.
About This Manual xvii
Chapter 8
Chapter 9
Describes how to configure a shared SCSI bus for tape drive,
tape loader, or tape library usage.
Contains information about setting up a shared SCSI bus, SCSI bus
requirements, and how to connect storage to a shared SCSI bus using
external termination or radial connections to non-UltraSCSI devices.
Describes how to prepare systems for a TruCluster Server configuration,
and how to connect host bus adapters to shared storage using external
termination or radial connection to non-UltraSCSI devices.
Chapter 10
Related Documents
Users of the TruCluster Server product can consult the following manuals for
assistance in cluster installation, administration, and programming tasks:
•
TruCluster Server Software Product Description (SPD) — The
comprehensive description of the TruCluster Server Version 5.0A
product. You can find the latest version of the SPD and other TruCluster
Server documentation at the following URL:
http://www.unix.digital.com/faqs/publications/pub_page/cluster_list.html
•
•
•
•
•
Release Notes — Provides important information about TruCluster
Server Version 5.0A.
Technical Overview — Provides an overview of the TruCluster Server
technology.
Software Installation — Describes how to install the TruCluster Server
product.
Cluster Administration — Describes cluster-specific administration
tasks.
Highly Available Applications — Describes how to deploy applications on
a TruCluster Server cluster.
The UltraSCSI Configuration Guidelines document provides guidelines
regarding UltraSCSI configurations.
For information about setting up a RAID subsystem, see the following
documentation as appropriate for your configuration:
•
•
•
•
DEC RAID Subsystem User ’s Guide
HS Family of Array Controllers User ’s Guide
RAID Array 310 Configuration and Maintenance Guide User ’s Guide
Configuring Your StorageWorks Subsystem HSZ40 Array Controllers
HSOF Version 3.0
•
Getting Started RAID Array 450 V5.4 for Compaq Tru64 UNIX
Installation Guide
xviii About This Manual
•
•
•
HSZ70 Array Controller HSOF Version 7.0 Configuration Manual
HSZ80 Array Controller ACS Version 8.2
Compaq StorageWorks HSG80 Array Controller ACS Version 8.5
Configuration Guide
•
Compaq StorageWorks HSG80 Array Controller ACS Version 8.5 CLI
Reference Guide
•
Wwidmgr User ’s Manual
For information about the tape devices, see the following documentation:
•
•
•
•
•
•
•
•
•
TZ88 DLT Series Tape Drive Owner ’s Manual
TZ89 DLT Series Tape Drive User ’s Guide
TZ885 Model 100/ 200 GB DLT 5-Cartridge MiniLibrary Owner ’s Manual
TZ887 Model 140/ 280 GB DLT 7-Cartridge MiniLibrary Owner ’s Manual
TL881 MiniLibrary System User ’s Guide
TL881 MiniLibrary Drive Upgrade Procedure
Pass-Through Expansion Kit Installation Instructions
TL891 MiniLibrary System User ’s Guide
TL81X/ TL894 Automated Tape Library for DLT Cartridges Facilities
Planning and Installation Guide
•
TL81X/ TL894 Automated Tape Library for DLT Cartridges Diagnostic
Software User ’s Manual
•
•
•
•
•
TL895 DLT Tape Library Facilities Planning and Installation Guide
TL895 DLT Library Operator ’s Guide
TL895 DLT Tape Library Diagnostic Software User ’s Manual
TL895 Drive Upgrade Instructions
TL82X/ TL893/ TL896 Automated Tape Library for DLT Cartridges
Facilities Planning and Installation Guide
•
•
TL82X/ TL893/ TL896 Automated Tape Library for DLT Cartridges
Operator ’s Guide
TL82X/ TL893/ TL896 Automated Tape Library for DLT Cartridges
Diagnostic Software User ’s Manual
•
•
TL82X Cabinet-to-Cabinet Mounting Instructions
TL82X/ TL89X MUML to MUSL Upgrade Instructions
The Golden Eggs Visual Configuration Guide provides configuration
diagrams of workstations, servers, storage components, and clustered
About This Manual xix
systems. It is available on line in PostScript and Portable Document Format
(PDF) formats at:
http://www.compaq.com/info/golden-eggs
At this URL you will find links to individual system, storage, or cluster
configurations. You can order the document through the Compaq Literature
Order System (LOS) as order number EC-R026B-36.
In addition, you should have available the following manuals from the Tru64
UNIX documentation set:
•
•
•
•
Installation Guide
Release Notes
System Administration
Network Administration
You should also have the hardware documentation for the systems, SCSI
controllers, disk storage shelves or RAID controllers, and any other
hardware you plan to install.
Documentation for the following optional software products will be useful if
you intend to use these products with TruCluster Server:
•
•
•
•
•
•
Compaq Analyze (DS20 and ES40)
DECevent™ (AlphaServers other than the DS20 and ES40)
Logical Storage Manager (LSM)
NetWorker
Advanced File System (AdvFS) Utilities
Performance Manager
Reader’s Comments
Compaq welcomes any comments and suggestions you have on this and
other Tru64 UNIX manuals.
You can send your comments in the following ways:
•
•
Fax: 603-884-0120 Attn: UBPG Publications, ZKO3-3/Y32
Internet electronic mail: [email protected]
A Reader ’s Comment form is located on your system in the following
location:
/usr/doc/readers_comment.txt
•
Mail:
Compaq Computer Corporation
xx About This Manual
UBPG Publications Manager
ZKO3-3/Y32
110 Spit Brook Road
Nashua, NH 03062-2698
A Reader ’s Comment form is located in the back of each printed manual.
The form is postage paid if you mail it in the United States.
Please include the following information along with your comments:
•
The full title of the book and the order number. (The order number is
printed on the title page of this book and on its back cover.)
•
The section numbers and page numbers of the information on which
you are commenting.
•
•
The version of Tru64 UNIX that you are using.
If known, the type of processor that is running the Tru64 UNIX software.
The Tru64 UNIX Publications group cannot respond to system problems
or technical support inquiries. Please address technical questions to your
local system vendor or to the appropriate Compaq technical support office.
Information provided with the software media explains how to send problem
reports to Compaq.
Conventions
The following typographical conventions are used in this manual:
#
A number sign represents the superuser prompt.
% cat
Boldface type in interactive examples indicates
typed user input.
file
Italic (slanted) type indicates variable values,
placeholders, and function argument names.
.
.
.
A vertical ellipsis indicates that a portion of an
example that would normally be present is not
shown.
cat(1)
A cross-reference to a reference page includes
the appropriate section number in parentheses.
For example, cat(1) indicates that you can find
information on the cat command in Section 1 of
the reference pages.
About This Manual xxi
Bold text indicates a term that is defined in the
glossary.
clu ster
xxii About This Manual
1
Introduction
This chapter introduces the TruCluster Server product and some basic
cluster hardware configuration concepts.
Subsequent chapters describe how to set up and maintain TruCluster Server
hardware configurations. See the TruCluster Server Software Installation
manual for information about software installation; see the TruCluster
Server Cluster Administration manual for detailed information about setting
up member systems and highly available applications.
1.1 The TruCluster Server Product
TruCluster Server, the newest addition to the Compaq Tru64 UNIX
TruCluster Software products family, extends single-system management
capabilities to clusters. It provides a clusterwide namespace for files and
directories, including a single root file system that all cluster members
share. It also offers a cluster alias for the Internet protocol suite (TCP/IP) so
that a cluster appears as a single system to its network clients.
TruCluster Server preserves the availability and performance features found
in the earlier TruCluster products:
•
Like the TruCluster Available Server Software and TruCluster
Production Server products, TruCluster Server lets you deploy highly
available applications that have no embedded knowledge that they are
executing in a cluster. They can access their disk data from any member
in the cluster.
•
Like the TruCluster Production Server Software product, TruCluster
Server lets you run components of distributed applications in parallel,
providing high availability while taking advantage of cluster-specific
synchronization mechanisms and performance optimizations.
TruCluster Server augments the feature set of its predecessors by allowing
all cluster members access to all file systems and all storage in the cluster,
regardless of where they reside. From the viewpoint of clients, a TruCluster
Server cluster appears to be a single system; from the viewpoint of a system
administrator, a TruCluster Server cluster is managed as if it were a single
system. Because TruCluster Server has no built-in dependencies on the
architectures or protocols of its private cluster interconnect or shared storage
Introduction 1–1
interconnect, you can more easily alter or expand your cluster ’s hardware
configuration as newer and faster technologies become available.
1.2 Overview of the TruCluster Server Hardware
Configuration
A TruCluster Server hardware configuration consists of a number of highly
specific hardware components:
•
•
TruCluster Server currently supports from one to eight member systems.
There must be sufficient internal and external SCSI controllers, Fibre
Channel host bus adapters, and disks to provide sufficient storage for
the applications.
•
The clusterwide root (/), /usr, and /var file systems should be on
a shared SCSI bus. We recommend placing all member system boot
disks on a shared SCSI bus. If you have a quorum disk, it must be on
a shared SCSI bus.
_____________________ Note _____________________
The clusterwide root (/), /usr, and /var file systems, the
member system boot disks, and the quorum disk may be
located behind a RAID array controller, including the HSG80
controller (Fibre Channel).
•
You need to allocate a number of Internet Protocol (IP) addresses from
one IP subnet to allow client access to the cluster. The IP subnet has
to be visible to the clients directly or through routers. The miminum
number of allocated addresses is equal to the number of cluster member
systems plus one (for the cluster alias), depending on the type of cluster
alias configuration.
For client access, TruCluster Server allows you to configure any number
of monitored network adapters (using a redundant array of independent
network adapters (NetRAIN) and Network Interface Failure Finder
(NIFF) facilities of the Tru64 UNIX operating system).
•
TruCluster Server requires at least one peripheral component
interconnect (PCI) Memory Channel adapter on each system. The
Memory Channel adapters comprise the cluster interconnect for
TruCluster Server, providing host-to-host communications. For a cluster
with two systems, a Memory Channel hub is optional; the Memory
Channel adapters can be connected with a cable.
If there are more than two systems in the cluster, a Memory Channel
hub is required. The Memory Channel hub is a PC-class enclosure that
1–2 Introduction
contains up to eight linecards. The Memory Channel adapter in each
system in the cluster is connected to the Memory Channel hub.
One or two Memory Channel adapters can be used with TruCluster
Server. When dual Memory Channel adapters are installed, if the
Memory Channel adapter being used for cluster communication fails, the
communication will fail over to the other Memory Channel.
1.3 Memory Requirements
Cluster members require a minimum of 128 MB of memory.
1.4 Minimum Disk Requirements
This section provides an overview of the minimum file system or disk
requirements for a two-node cluster. For more information on the amount
of space required for each required cluster file system, see the TruCluster
Server Software Installation manual.
1.4.1 Disks Needed for Installation
You need to allocate disks for the following uses:
•
One or more disks to hold the Tru64 UNIX operating system. The disk(s)
are either private disk(s) on the system that will become the first cluster
member, or disk(s) on a shared bus that the system can access.
•
•
•
One or more disks on a shared SCSI bus to hold the clusterwide root (/),
/usr, and /var AdvFS file systems.
One disk per member, normally on a shared SCSI bus, to hold member
boot partitions.
Optionally, one disk on a shared SCSI bus to act as the quorum disk. See
Section 1.4.1.4, and for a more detailed discussion of the quorum disk,
see the TruCluster Server Cluster Administration manual.
The following sections provide more information about these disks.
Figure 1–1 shows a generic two-member cluster with the required file
systems.
1.4.1.1 Tru64 UNIX Operating System Disk
The Tru64 UNIX operating system is installed using AdvFS file systems on
one or more disks on the system that will become the first cluster member.
For example:
dsk0a
dsk0g
dsk0h
root_domain#root
usr_domain#usr
var_domain#var
Introduction 1–3
The operating system disk (Tru64 UNIX disk) cannot be used as a
clusterwide disk, a member boot disk, or as the quorum disk.
Because the Tru64 UNIX operating system will be available on the first
cluster member, in an emergency, after shutting down the cluster, you have
the option of booting the Tru64 UNIX operating system and attempting to
fix the problem. See the TruCluster Server Cluster Administration manual
for more information.
1.4.1.2 Clusterwide Disk(s)
When you create a cluster, the installation scripts copy the Tru64 UNIX
root (/), /usr, and /var file systems from the Tru64 UNIX disk to the disk
or disks you specify.
We recommend that the disk or disks used for the clusterwide file systems
be placed on a shared SCSI bus so that all cluster members have access to
these disks.
During the installation, you supply the disk device names and partitions
that will contain the clusterwide root (/), /usr, and /var file systems. For
example, dsk3b, dsk4c, and dsk3g:
dsk3b
dsk4c
dsk3g
cluster_root#root
cluster_usr#usr
cluster_var#var
The /var fileset cannot share the cluster_usr domain, but must be a
separate domain, cluster_var. Each AdvFS file system must be a separate
partition; the partitions do not have to be on the same disk.
If any partition on a disk is used by a clusterwide file system, only
clusterwide file systems can be on that disk. A disk containing a clusterwide
file system cannot also be used as the member boot disk or as the quorum
disk.
1.4.1.3 Member Boot Disk
Each member has a boot disk. A boot disk contains that member ’s boot,
swap, and cluster-status partitions. For example, dsk1 is the boot disk for
the first member and dsk2 is the boot disk for the second member:
dsk1
dsk2
first member’s boot disk [pepicelli]
second member’s boot disk [polishham]
The installation scripts reformat each member ’s boot disk to contain three
partitions: an a partition for that member ’s root (/) file system, a b partition
for swap, and an h partition for cluster status information. (There are no
/usr or /var file systems on a member ’s boot disk.)
1–4 Introduction
A member boot disk cannot contain one of the clusterwide root (/), /usr,
and /var file systems. Also, a member boot disk cannot be used as the
quorum disk. A member disk can contain more than the three required
partitions. You can move the swap partition off the member boot disk. See
the TruCluster Server Cluster Administration manual for more information.
1.4.1.4 Quorum Disk
The quorum disk allows greater availability for clusters consisting of two
members. Its h partition contains cluster status and quorum information.
See the TruCluster Server Cluster Administration manual for a discussion of
how and when to use a quorum disk.
The following restrictions apply to the use of a quorum disk:
•
•
A cluster can have only one quorum disk.
The quorum disk should be on a shared bus to which all cluster members
are directly connected. If it is not, members that do not have a direct
connection to the quorum disk may lose quorum before members that
do have a direct connection to it.
•
The quorum disk must not contain any data. The clu_quorum command
will overwrite existing data when initializing the quorum disk. The
integrity of data (or file system metadata) placed on the quorum disk
from a running cluster is not guaranteed across member failures.
This means that the member boot disks and the disk holding the
clusterwide root (/) cannot be used as quorum disks.
•
•
The quorum disk can be small. The cluster subsystems use only 1 MB
of the disk.
A quorum disk can have either 1 vote or no votes. In general, a quorum
disk should always be assigned a vote. You might assign an existing
quorum disk no votes in certain testing or transitory configurations,
such as a one-member cluster (in which a voting quorum disk introduces
a second point of failure).
•
You cannot use the Logical Storage Manager (LSM) on the quorum disk.
1.5 Generic Two-Node Cluster
This section describes a generic two-node cluster with the minimum disk
layout of four disks. Note that additional disks may be needed for highly
available applications. In this section, and the following sections, the type
of PCI SCSI bus adapter is not significant. Also, although an important
consideration, SCSI bus cabling, including Y cables or trilink connectors,
termination, and the use of UltraSCSI hubs is not considered at this time.
Introduction 1–5
Figure 1–1 shows a generic two-node cluster with the minimum number
of disks.
•
•
•
•
Tru64 UNIX disk
Clusterwide root (/), /usr, and /var
Member 1 boot disk
Member 2 boot disk
A minimum configuration cluster may have reduced availability due to the
lack of a quorum disk. As shown, with only two-member systems, both
systems must be operational to achieve quorum and form a cluster. If only
one system is operational, it will loop, waiting for the second system to boot
before a cluster can be formed. If one system crashes, you lose the cluster.
Figure 1–1: Two-Node Cluster with Minimum Disk Configuration and No
Quorum Disk
Network
Member
System
1
Member
System
2
Memory Channel
PCI SCSI
Adapter
PCI SCSI
Adapter
Tru64
UNIX
Disk
Shared SCSI Bus
Cluster File
System
root (/)
/usr
Member 1
Member 2
root (/)
swap
root (/)
swap
/var
ZK-1587U-AI
Figure 1–2 shows the same generic two-node cluster as shown in Figure 1–1,
but with the addition of a quorum disk. By adding a quorum disk, a cluster
may be formed if both systems are operational, or if either of the systems
and the quorum disk is operational. This cluster has a higher availability
than the cluster shown in Figure 1–1. See the TruCluster Server Cluster
1–6 Introduction
Administration manual for a discussion of how and when to use a quorum
disk.
Figure 1–2: Generic Two-Node Cluster with Minimum Disk Configuration
and Quorum Disk
Network
Member
System
1
Member
System
2
Memory Channel
PCI SCSI
Adapter
PCI SCSI
Adapter
Tru64
UNIX
Disk
Shared SCSI Bus
Cluster File
System
root (/)
/usr
Member 1
Member 1
Quorum
root (/)
swap
root (/)
swap
/var
ZK-1588U-AI
1.6 Growing a Cluster from Minimum Storage to a NSPOF
Cluster
The following sections take a progression of clusters from a cluster with
minimum storage to a no-single-point-of-failure (NSPOF) cluster; a cluster
where one hardware failure will not interrupt the cluster operation:
•
•
•
A cluster with minimum storage for highly available applications
(Section 1.6.1).
A cluster with more storage, but the single SCSI bus is a single point
of failure (Section 1.6.2).
Adding a second SCSI bus allows the use of LSM to mirror the /usr and
/var file systems and data disks. However, as LSM cannot mirror the
root (/), member system boot, swap, or quorum disks, so full redundancy
is not achieved (Section 1.6.3).
Introduction 1–7
•
•
Using a RAID array controller in transparent failover mode allows the
use of hardware RAID to mirror the disks. However, without a second
SCSI bus, second Memory Channel, and redundant networks, this
configuration is still not a NSPOF cluster (Section 1.6.4).
By using an HSZ70, HSZ80, or HSG80 with multiple-bus failover enabled
you can use two shared SCSI buses to access the storage. Hardware
RAID is used to mirror the root (/), /usr, and /var file systems, and
the member system boot disks, data disks, and quorum disk (if used).
A second Memory Channel, redundant networks, and redundant power
must also be installed to achieve a NSPOF cluster (Section 1.6.5).
1.6.1 Two-Node Clusters Using an UltraSCSI BA356 Storage Shelf
and Minimum Disk Configurations
This section takes the generic illustrations of our cluster example one step
further by depicting the required storage in storage shelves. The storage
shelves could be BA350, BA356 (non-UltraSCSI), or UltraSCSI BA356s.
The BA350 is the oldest model, and can only respond to SCSI IDs 0-6. The
non-Ultra BA356 can respond to SCSI IDs 0-6 or 8-14 (see Section 3.2). The
UltraSCSI BA356 also responds to SCSI IDs 0-6 or 8-14, but also can operate
at UltraSCSI speeds (see Section 3.2).
Figure 1–3 shows a TruCluster Server configuration using an UltraSCSI
BA356 storage unit. The DS-BA35X-DA personality module used in the
UltraSCSI BA356 storage unit is a differential-to-single-ended signal
converter, and therefore accepts differential inputs.
______________________
Note _______________________
The figures in this section are generic drawings and do not show
shared SCSI bus termination, cable names, and so forth.
1–8 Introduction
Figure 1–3: Minimum Two-Node Cluster with UltraSCSI BA356 Storage Unit
Network
Member
System
1
Member
System
2
Memory
Channel
Interface
Memory Channel
Memory Channel
Host Bus Adapter (ID 6)
Host Bus Adapter (ID 7)
UltraSCSI
BA356
Shared
SCSI
Bus
Shared
SCSI
Bus
Tru64
UNIX
Disk
Clusterwide
/, /usr, /var
ID 0
DS-BA35X-DA
Personality
Module
Member 1
Boot Disk
ID 1
ID 2
ID 3
Member 2
Boot Disk
Quorum
Disk
ID 4
ID 5
Clusterwide
Data Disks
Do not use for
data disk. May
be used for
ID 6
redundant power
supply.
PWR
ZK-1591U-AI
The configuration shown in Figure 1–3 might represent a typical small or
training configuration with TruCluster Server Version 5.0A required disks.
In this configuration, because of the TruCluster Server Version 5.0A disk
requirements, there will only be two disks available for highly available
applications.
______________________
Note _______________________
Slot 6 in the UltraSCSI BA356 is not available because SCSI ID 6
is generally used for a member system SCSI adapter. However,
Introduction 1–9
this slot can be used for a second power supply to provide fully
redundant power to the storage shelf.
Note that with the use of the cluster file system (See the TruCluster Server
Cluster Administration manual for a discussion of the cluster file system),
the clusterwide root (/), /usr, and /var file systems could be physically
placed on a private bus of either of the member systems. But, if that member
system was not available, the other member system(s) would not have access
to the clusterwide file systems. Therefore, placing the clusterwide root (/),
/usr, and /var file systems on a private bus is not recommended.
Likewise, the quorum disk could be placed on the local bus of either of the
member systems. If that member was not available, quorum could never be
reached in a two-node cluster. Placing the quorum disk on the local bus of a
member system is not recommended as it creates a single point of failure.
The individual member boot and swap partitions could also be placed on
a local bus of either of the member systems. If the boot disk for member
system 1 was on a SCSI bus internal to member 1, and the system was
unavailable due to a boot disk problem, other systems in the cluster could
not access the disk for possible repair. If the member system boot disks are
on a shared SCSI bus, they can be accessed by other systems on the shared
SCSI bus for possible repair.
By placing the swap partition on a system’s internal SCSI bus, you reduce
total traffic on the shared SCSI bus by an amount equal to the system’s
swap volume.
TruCluster Server Version 5.0A configurations require one or more disks to
hold the Tru64 UNIX operating system. The disk(s) are either private disk(s)
on the system that will become the first cluster member, or disk(s) on a
shared bus that the system can access.
We recommend that you place the /usr, /var, member boot disks, and
quorum disk on a shared SCSI bus connected to all member systems. After
installation, you have the option to reconfigure swap and can place the swap
disks on an internal SCSI bus to increase performance. See the TruCluster
Server Cluster Administration manual for more information.
1.6.2 Two-Node Clusters Using UltraSCSI BA356 Storage Units with
Increased Disk Configurations
The configuration shown in Figure 1–3 is a minimal configuration, with a
lack of disk space for highly available applications. Starting with Tru64
UNIX Version 5.0, 16 devices are supported on a SCSI bus. Therefore,
1–10 Introduction
multiple BA356 storage units can be used on the same SCSI bus to allow
more devices on the same bus.
Figure 1–4 shows the configuration in Figure 1–3 with a second UltraSCSI
BA356 storage unit that provides an additional seven disks for highly
available applications.
Figure 1–4: Two-Node Cluster with Two UltraSCSI DS-BA356 Storage Units
Network
Member
System
1
Member
System
2
Memory
Channel
Interface
Memory Channel
Memory Channel
Host Bus Adapter (ID 6)
Host Bus Adapter (ID 7)
UltraSCSI
BA356
UltraSCSI
BA356
Shared
SCSI
Bus
Tru64
UNIX
Disk
Clusterwide
/, /usr, /var
ID 8
ID 9
ID 0
Member 1
Boot Disk
ID 1
ID 2
Member 2
Boot Disk
ID 10
Quorum
Disk
Data
ID 11
Disks
ID 3
ID 4
ID 5
ID 4
ID 5
ID 12
Data
disks
ID 13
Do not use for
data disk. May
be used for
redundant power
supply.
ID 14 or
redundant
power
ID 6
supply
PWR
PWR
ZK-1590U-AI
This configuration, while providing more storage, has a single SCSI bus that
presents a single point of failure. Providing a second SCSI bus would allow
the use of the Logical Storage Manager (LSM) to mirror the /usr and /var
file systems and the data disks across SCSI buses, removing the single SCSI
bus as a single point of failure for these file systems.
Introduction 1–11
1.6.3 Two-Node Configurations with UltraSCSI BA356 Storage Units
and Dual SCSI Buses
By adding a second shared SCSI bus, you now have the capability to use the
Logical Storage Manager (LSM) to mirror data disks, and the clusterwide
/usr and /var file systems across SCSI buses.
______________________
Note _______________________
You cannot use LSM to mirror the clusterwide root (/), member
system boot, swap, or quorum disks, but you can use hardware
RAID.
Figure 1–5 shows a small cluster configuration with dual SCSI buses using
LSM to mirror the clusterwide /usr and /var file systems and the data
disks.
1–12 Introduction
Figure 1–5: Two-Node Configurations with UltraSCSI BA356 Storage Units
and Dual SCSI Buses
Network
Tru64
UNIX
Disk
Member
System
1
Member
System
2
Memory
Channel
Interface
Memory Channel
Memory Channel
Host Bus Adapter (ID 6)
Host Bus Adapter (ID 6)
Host Bus Adapter (ID 7)
Host Bus Adapter (ID 7)
UltraSCSI
BA356
UltraSCSI
BA356
UltraSCSI
BA356
UltraSCSI
BA356
Mirrored
/usr, /var
Mirrored
Clusterwide
/, /usr, /var
ID 8
Data Disk
ID 8
ID 0
ID 1
ID 2
ID 3
ID 0
ID 1
Data Disk
Member 1
Boot Disk
Mirrored
Data Disk
ID 9
Data Disk
ID 9
Not Used
Not Used
Member 2
Boot Disk
ID 10
Data Disk
Mirrored
Data Disk
ID 10
ID 11
ID 12
ID 13
ID 2
ID 3
ID 4
ID 5
Mirrored
Data Disk
ID 11
Data Disk
Quorum
Disk
Not Used
Mirrored
Data Disk
ID 12
Data Disk
Mirrored
Data Disk
Data Disk
Data Disk
ID 4
ID 5
Mirrored
Data Disk
Mirrored
Data Disk
ID 13
Data Disk
Redundant
PWR or not
used
Redundant
PWR or not
used
ID 14 or
PWR
ID 14 or
PWR
ID 6
ID 6
PWR
PWR
PWR
PWR
ZK-1593U-AI
By using LSM to mirror the /usr and /var file systems and the data disks,
we have achieved higher availability. But, even if you have a second Memory
Channel and redundant networks, because we cannot use LSM to mirror the
clusterwide root (/), quorum, or the member boot disks, we do not have a
no-single-point-of-failure (NSPOF) cluster.
1.6.4 Using Hardware RAID to Mirror the Clusterwide Root File
System and Member System Boot Disks
You can use hardware RAID with any of the supported RAID array
controllers to mirror the clusterwide root (/), quorum, and member boot
disks. Figure 1–6 shows a cluster configuration using an HSZ70 RAID array
controller. An HSZ40, HSZ50, HSZ80, or HSG80 could be used instead of the
Introduction 1–13
HSZ70. The array controllers can be configured as a dual redundant pair. If
you want the capability to fail over from one controller to another controller,
you must install the second controller. Also, you must set the failover mode.
Figure 1–6: Cluster Configuration with HSZ70 Controllers in Transparent
Failover Mode
Network
Member
System
1
Member
System
2
Memory
Channel
Interface
Memory Channel
Memory Channel
Host Bus Adapter (ID 6)
Host Bus Adapter (ID 7)
Tru64
UNIX
Disk
HSZ70
HSZ70
StorageWorks
RAID Array 7000
ZK-1589U-AI
In Figure 1–6 the HSZ40, HSZ50, HSZ70, HSZ80, or HSG80 has transparent
failover mode enabled (SET FAILOVER COPY = THIS_CONTROLLER). In
transparent failover mode, both controllers are connected to the same shared
SCSI bus and device buses. Both controllers service the entire group of
storagesets, single-disk units, or other storage devices. Either controller can
continue to service all of the units if the other controller fails.
______________________
Note _______________________
The assignment of HSZ target IDs can be balanced between the
controllers to provide better system performance. See the RAID
array controller documentation for information on setting up
storagesets.
1–14 Introduction
Note that in the configuration shown in Figure 1–6, there is only one shared
SCSI bus. Even by mirroring the clusterwide root and member boot disks,
the single shared SCSI bus is a single point of failure.
1.6.5 Creating a NSPOF Cluster
To create a no-single-point-of-failure (NSPOF) cluster:
•
Use hardware RAID to mirror the clusterwide root (/), /usr, and /var
file systems, the member boot disks, quorum disk (if present), and data
disks
•
Use at least two shared SCSI buses to access dual-redundant RAID
array controllers set up for multiple-bus failover mode (HSZ70, HSZ80,
and HSG80)
•
•
•
•
Install a second Memory Channel interface for redundancy
Install redundant power supplies
Install redundant networks
Connect the systems and storage to an uninterruptable power supply
(UPS)
Tru64 UNIX support for multipathing provides support for multiple-bus
failover.
______________________ Notes ______________________
Only the HSZ70, HSZ80, and HSG80 are capable of supporting
multiple-bus failover (SET MULTIBUS_FAILOVER COPY =
THIS_CONTROLLER).
Partitioned storagesets and partitioned single-disk units cannot
function in multiple-bus failover dual-redundant configurations
with the HSZ70 or HSZ80. You must delete any partitions before
configuring the controllers for multiple-bus failover.
Partitioned storagesets and partitioned single-disk units are
supported with the HSG80 and ACS V8.5.
Figure 1–7 shows a cluster configuration with dual-shared SCSI buses and a
storage array with dual-redundant HSZ70s. If there is a failure in one SCSI
bus, the member systems can access the disks over the other SCSI bus.
Introduction 1–15
Figure 1–7: NSPOF Cluster using HSZ70s in Multiple-Bus Failover Mode
Networks
Tru64
UNIX
Disk
Memory
Channel
Interfaces
Member System 1
Member System 2
Memory Channel (mca1)
Memory Channel (mca1)
Memory Channel (mca0)
Memory Channel (mca0)
Host Bus Adapter (ID 6)
Host Bus Adapter (ID 6)
Host Bus Adapter (ID 7)
Host Bus Adapter (ID 7)
HSZ70
HSZ70
StorageWorks
RAID Array 7000
ZK-1594U-AI
Figure 1–8 shows a cluster configuration with dual-shared Fibre Channel
SCSI buses and a storage array with dual-redundant HSG80s configured for
multiple-bus failover.
1–16 Introduction
Figure 1–8: NSPOF Fibre Channel Cluster using HSG80s in Multiple-Bus
Failover Mode
Member
System
1
Member
System
2
KGPSA
KGPSA
KGPSA
KGPSA
DSGGA
DSGGA
RA8000/
ESA12000
RA8000/
ESA12000
ZK-1533U-AI
1.7 Overview of Setting up the TruCluster Server Hardware
Configuration
To set up a TruCluster Server hardware configuration, follow these steps:
1. Plan your hardware configuration. (See Chapter 3, Chapter 4,
Chapter 6, Chapter 9, and Chapter 10).
2. Draw a diagram of your configuration.
3. Compare your diagram with the examples in Chapter 3, Chapter 6,
and Chapter 9.
4. Identify all devices, cables, SCSI adapters, and so forth. Use the
diagram you just constructed.
5. Prepare the shared storage by installing disks and configuring any RAID
controller subsystems (See Chapter 3, Chapter 6, and Chapter 9 and the
documentation for the StorageWorks enclosure or RAID controller).
Introduction 1–17
6. Install signal converters in the StorageWorks enclosures, if applicable
(see Chapter 3 and Chapter 9).
7. Connect storage to the shared SCSI buses. Terminate each bus. Use
Y cables or trilink connectors where necessary (see Chapter 3 and
Chapter 9).
For a Fibre Channel configuration, connect the HSG80 controllers to
the switches. You want the HSG80 to recognize the connections to the
systems when the systems are powered on.
8. Prepare the member systems by installing:
•
Additional Ethernet or Asynchronous Transfer Mode (ATM) network
adapters for client networks (see Chapter 7).
•
SCSI bus adapters. Ensure that adapter terminators are set
correctly. Connect the systems to the shared SCSI bus (see
Chapter 4 or Chapter 10).
•
•
The KGPSA host bus adapter for Fibre Channel configurations.
Ensure that the KGPSA is operating in the correct mode (FABRIC or
LOOP). Connect the KGPSA to the switch (see Chapter 6).
Memory Channel adapters. Ensure that jumpers are set correctly
(see Chapter 5).
9. Connect the Memory Channel adapters to each other or to the Memory
Channel hub as appropriate (see Chapter 5).
10. Turn on the Memory Channel hubs and storage shelves, then turn on
the member systems.
11. Install the firmware, set SCSI IDs, and enable fast bus speed as
necessary (see Chapter 4 and Chapter 10).
12. Display configuration information for each member system, and ensure
that all shared disks are seen at the same device number (see Chapter 4,
Chapter 6, or Chapter 10).
1–18 Introduction
2
Hardware Requirements and Restrictions
This chapter describes the hardware requirements and restrictions for
a TruCluster Server cluster. It includes lists of supported cables, trilink
connectors, Y cables, and terminators.
See the TruCluster Server Software Product Description (SPD) for the latest
information about supported hardware.
2.1 TruCluster Server Member System Requirements
The requirements for member systems in a TruCluster Server cluster are as
follows:
•
Each supported member system requires a minimum firmware revision.
See the Release Notes Overview supplied with the Alpha Systems
Firmware Update CD-ROM.
You can also obtain firmware information from the Web at
http://www.compaq.com/support/. Select Alpha Systems from
the downloadable drivers & utilities menu. Check the release
notes for the appropriate system type to determine the firmware version
required.
•
TruCluster Server Version 5.0A supports eight-member cluster
configurations as follows:
–
Fibre Channel: Eight-member systems may be connected to common
storage over Fibre Channel in a fabric (switch) configuration.
–
Parallel SCSI: Only four of the member systems may be connected to
any one SCSI bus, but you can have multiple SCSI buses connected
to different sets of nodes, and the sets of nodes may overlap. We
recommend you use a DS-DWZZH-05 UltraSCSI hub with fair
arbitration enabled when connecting four-member systems to a
common SCSI bus.
•
TruCluster Server does not support the XMI CIXCD on an AlphaServer
8x00, GS60, GS60E, or GS140 system.
2.2 Memory Channel Restrictions
The Memory Channel interconnect is used for cluster communications
between the member systems.
Hardware Requirements and Restrictions 2–1
There are currently three versions of the Memory Channel product; Memory
Channel 1, Memory Channel 1.5, and Memory Channel 2. The Memory
Channel 1 and Memory Channel 1.5 products are very similar (the PCI
adapter for both versions is the CCMAA module) and are generally referred
to as MC1 throughout this manual. The Memory Channel 2 product
(CCMAB module) is referred to as MC2.
Ensure that you abide by the following Memory Channel restrictions:
•
•
The DS10, DS20, DS20E, and ES40 systems only support MC2 hardware.
If redundant Memory Channel adapters are used with a DS10, they must
be jumpered for 128 MB and not the default of 512 MB.
•
If you use the MC API library functions in a 2-node TruCluster Server
configuration, you cannot use virtual hub mode, you must use a Memory
Channel hub and standard hub mode.
•
•
If you have redundant MC2 modules on a system jumpered for 512 MB,
you cannot have two MC2 modules on the same PCI bus.
The MC1 adapter cannot be cabled to a MC2 adapter.
–
–
Do not use the BC12N link cable with the CCMAB MC2 adapter.
Do not use the BN39B link cable with the CCMAA MC1 adapter.
•
•
Redundant Memory Channels are supported within a mixed Memory
Channel configuration, as long as MC1 adapters are connected to other
MC1 adapters and MC2 adapters are connected to MC2 adapters.
A Memory Channel interconnect can use either virtual hub mode (two
member systems connected without a Memory Channel hub) or standard
mode (two or more systems connected to a hub). A TruCluster Server
cluster with three or more member systems must be jumpered for
standard hub mode and requires a Memory Channel hub.
•
If Memory Channel modules are jumpered for virtual hub mode, all
Memory Channel modules on a system must be jumpered in the same
manner, either virtual hub 0 (VH0) or virtual hub 1 (VH1). You cannot
have one Memory Channel module jumpered for VH0 and another
jumpered for VH1 on the same system.
•
•
The maximum length of an MC1 BC12N and MC2 BN39B link cable is
10 meters.
Always check a Memory Channel link cable for bent or broken pins.
Be sure that you do not bend or break any pins when you connect or
disconnect a cable.
•
For AlphaServer 1000A systems, the Memory Channel adapter must be
installed on the primary PCI (in front of the PCI-to-PCI bridge chip) in
PCI slots 11, 12, or 13 (the top three slots).
2–2 Hardware Requirements and Restrictions
•
For AlphaServer 2000 systems, the B2111-AA module must be at
Revision H or higher.
For AlphaServer 2100 systems, the B2110-AA module must be at
Revision L or higher.
Use the examine console command to determine if these modules are
at a supported revision as follows:
P00>>> examine -b econfig:20008
econfig: 20008 04
P00>>>
If a hexadecimal value of 04 or greater is returned, the I/O module
supports Memory Channel.
If a hexadecimal value of less than 04 is returned, the I/O module is not
supported for Memory Channel usage.
Order an H3095-AA module to upgrade an AlphaServer 2000 or an
H3096-AA module to upgrade an AlphaServer 2100 to support Memory
Channel.
•
•
•
For AlphaServer 2100A systems, the Memory Channel adapter must
be installed in PCI 4 through PCI 7 (slots 6, 7, 8, and 9), the bottom
four PCI slots.
For AlphaServer 8200, 8400, GS60, GS60E, or GS140 systems, the
Memory Channel adapter must be installed in slots 0-7 of a DWLPA
PCIA option; there are no restrictions for a DWLPB.
If a TruCluster Server cluster configuration utilizes multiple Memory
Channel adapters in standard hub mode, the Memory Channel adapters
must be connected to separate Memory Channel hubs. The first Memory
Channel adapter (mca0) in each system must be connected to one
Memory Channel hub. The second Memory Channel adapter (mcb0) in
each system must be connected to a second Memory Channel hub. Also,
each Memory Channel adapter on one system must be connected to the
same linecard in each Memory Channel hub.
2.3 Fibre Channel Requirements and Restrictions
Table 2–1 shows the supported AlphaServer systems with Fibre Channel
and the number of KGPSA-BC PCI-to-Fibre Channel adapters supported
on each system.
Hardware Requirements and Restrictions 2–3
Table 2–1: AlphaServer Systems Supported for Fibre Channel
AlphaServer
Number of KGPSA-BC Adapters
Supported
2
4
4
1
2
4
AlphaServer 800
AlphaServer 1200
AlphaServer 4000, 4000A, or 4100
Compaq AlphaServer DS10
Compaq AlphaServer DS20 and DS20E
Compaq AlphaServer ES40
AlphaServer 8200 or 8400a
32 (2 per DWLPB for throughput, 4
per DWLPB for connectivity)
Compaq AlphaServer GS60, GS60E,
32 (2 per DWLPB for throughput, 4
per DWLPB for connectivity)
and GS140a
a
The KGPSA-BC/CA PCI-to-Fibre Channel adapters are only supported on the DWLPB PCIA option;
they are not supported on the DWLPA.
The following requirements and restrictions apply to the use of Fibre
Channel with the TruCluster Server:
•
•
•
The HSG80 requires Array Control Software (ACS) Version 8.5.
A maximum of four member systems is supported.
The only supported Fibre Channel adapters are the KGPSA-BC and
KGPSA-CA PCI-to-Fibre Channel host bus adapters.
•
•
The only Fibre Channel switches supported are the 8/16 Port DSGGA or
DSGGB Fibre Channel switches.
The Fibre Channel switches support both shortwave (GBIC-SW) and
longwave (GBIC-LW) Giga Bit Interface Converter (GBIC) modules.
The GBIC-SW module supports 50-micron, multimode fibre cables with
the standard subscriber connector (SC connector) in lengths up to 500
meters. The GBIC-LW supports 9-micron, single-mode fibre cables with
the SC connector in lengths up to 10 kilometers.
The KGPSA-BC/CA PCI-to-Fibre Channel host bus adapters and the
HSG80 RAID controller support the 50-micron Gigabit Link Module
(GLM) for fibre connections. Therefore, only the 50-micron multimode
fibre optical cable is supported between the KGPSA and switch and the
switch and HSG80 for cluster configurations. You must install GBIC-SW
GBICs in the Fibre Channel switches for communication between the
switches and KGPSA or HSG80.
•
A maximum of three cascaded switches is supported, with a maximum of
two hops between switches. The maximum hop length is 10 km longwave
single-mode or 500 meters via shortwave multimode Fibre Channel cable.
2–4 Hardware Requirements and Restrictions
•
•
The Fibre Channel RAID Array 8000 (RA8000) midrange departmental
storage subsystem and Fibre Channel Enterprise Storage Array 12000
(ESA12000) house two HSG80 dual-channel controllers. There are
provisions for six UltraSCSI channels.
Only disk devices attached to the HSG80 Fibre Channel to Six Channel
UltraSCSI Array controller are supported with the TruCluster Server
product.
•
•
No tape devices are supported.
Tru64 UNIX Version 5.0A limits the number of Fibre Channel targets
to 126.
•
•
Tru64 UNIX Version 5.0A allows up to 255 LUNs per target.
The HSG80 supports transparent and multiple-bus failover mode when
used in a TruCluster Server Version 5.0A configuration. Multiple-bus
failover is recommended for high availability in a cluster.
•
•
A storage array with dual-redundant HSG80 controllers in transparent
mode failover is two targets and consumes four ports on a switch.
A storage array with dual-redundant HSG80 controllers in multiple-bus
failover is four targets and consumes 4 ports on a switch.
•
•
Each KGPSA is one target.
The HSG80 documentation refers to the controllers as Controllers A (top)
and B (bottom). Each controller provides two ports (left and right). (The
HSG80 documentation refers to these ports as Port 1 and 2, respectively.)
In transparent failover mode, only one left port and one right port are
active at any given time.
With transparent failover enabled, assuming that the left port of the top
controller and the right port of the bottom controller are active, if the top
controller fails in such a way that it can no longer properly communicate
with the switch, then its functions will automatically fail over to the
bottom controller (and vice versa).
•
•
In transparent failover mode, you can configure which controller
presents each HSG80 storage element (unit) to the cluster. Ordinarily,
the left port of either controller serves the units designated D0 through
D99, and the right port serves those designated D100 through D199.
In multiple-bus failover mode, all units (D0 through D199) are visible to
all host ports, but accessible only through one controller at any specific
time. The host can control the failover process by moving unit(s) from
one controller to the other controller.
Hardware Requirements and Restrictions 2–5
2.4 SCSI Bus Adapter Restrictions
To connect a member system to a shared SCSI bus, you must install a SCSI
bus adapter in an I/O bus slot.
The Tru64 UNIX operating system supports a maximum of 64 I/O buses.
TruCluster Server supports a total of 32 shared I/O buses using KZPSA-BB
host bus adapters, KZPBA-CB UltraSCSI host bus adapters, or KGPSA
Fibre Channel host bus adapters.
The following sections describe the SCSI adapter restrictions in more detail.
2.4.1 KZPSA-BB SCSI Adapter Restrictions
KZPSA-BB SCSI adapters have the following restrictions:
•
•
The KZPSA-BB requires A12 firmware.
If you have a KZPSA-BB adapter installed in an AlphaServer that
supports the bus_probe_algorithm console variable (for example, the
AlphaServer 800, 1000, 1000A, 2000, 2100, or 2100A systems support
the variable), you must set the bus_probe_algorithm console variable
to new by entering the following command:
>>> set bus_probe_algorithm new
Use the show bus_probe_algorithm console command to determine if
your system supports the variable. If the response is null or an error,
there is no support for the variable. If the response is anything other
than new, you must set it to new.
•
On AlphaServer 1000A and 2100A systems, updating the firmware on
the KZPSA-BB SCSI adapter is not supported when the adapter is
behind the PCI-to-PCI bridge.
2.4.2 KZPBA-CB SCSI Bus Adapter Restrictions
KZPBA-CB UltraSCSI adapters have the following restrictions:
•
Each system supporting the KZPBA-CB UltraSCSI host adapter limits
the number of adapters that may be installed. The maximum number of
KZPBA-CB UltraSCSI host adapters supported with TruCluster Server
follow:
–
–
–
AlphaServer 800: 2
AlphaServer 1000A and 1200: 4
AlphaServer 4000: 8; only one KZPBA-CB is supported in IOD0
(PCI0).
–
AlphaServer 4100: 5; only one KZPBA-CB is supported in IOD0
(PCI0).
2–6 Hardware Requirements and Restrictions
–
AlphaServer 8200, 8400, GS60, GS60E, GS140: 32
The KZPBA-CB is supported on the DWLPB only; it is not supported
on the DWLPA module.
–
–
–
AlphaServer DS10: 2
AlphaServer DS20/DS20E: 4
AlphaServer ES40: 5
•
•
•
A maximum of four HSZ50, HSZ70, or HSZ80 RAID array controllers can
be placed on a single KZPBA-CB UltraSCSI bus. Only two redundant
pairs of array controllers are allowed on one SCSI bus.
The KZPBA-CB requires ISP 1020/1040 firmware Version 5.57 (or
higher), which is available with the system SRM console firmware on the
Alpha Systems Firmware 5.3 Update CD-ROM (or later).
The maximum length of any differential SCSI bus segment is 25 meters,
including the length of the SCSI bus cables and SCSI bus internal to the
SCSI adapter, hub, or storage device. A SCSI bus may have more than
one SCSI bus segment (see Section 3.1).
•
See the KZPBA-CB UltraSCSI Storage Adapter Module Release Notes
(AA-R5XWD-TE) for more information.
2.5 Disk Device Restrictions
The restrictions for disk devices are as follows:
•
Disks on shared SCSI buses must be installed in external storage shelves
or behind a RAID array controller.
•
TruCluster Server does not support Prestoserve on any shared disk.
2.6 RAID Array Controller Restrictions
RAID array controllers provide high performance, high availability, and high
connectivity access to SCSI devices through a shared SCSI bus.
RAID controllers can be configured with the number of SCSI IDs as shown
in Table 2–2.
Table 2–2: RAID Controller SCSI IDs
RAID Controller
Number of SCSI IDs Supported
4
4
4
8
HSZ20
HSZ40
HSZ50
HSZ70
Hardware Requirements and Restrictions 2–7
Table 2–2: RAID Controller SCSI IDs (cont.)
RAID Controller
HSZ80
Number of SCSI IDs Supported
15
HSG80
N/A
2.7 SCSI Signal Converters
If you are using a standalone storage shelf with a single-ended SCSI
interface in your cluster configuration, you must connect it to a SCSI signal
converter. SCSI signal converters convert wide, differential SCSI to narrow
or wide, single-ended SCSI and vice versa. Some signal converters are
standalone desktop units and some are StorageWorks building blocks (SBBs)
that you install in storage shelves disk slots.
______________________
Note _______________________
The UltraSCSI hubs could probably be listed here as they contain
a DOC (DWZZA on a chip) chip, but they are covered separately
in Section 2.8.
The restrictions for SCSI signal converters are as follows:
•
If you remove the cover from a standalone unit, be sure to replace the
star washers on all four screws that hold the cover in place when you
reattach the cover. If the washers are not replaced, the SCSI signal
converter may not function correctly because of noise.
•
If you want to disconnect a SCSI signal converter from a shared SCSI
bus, you must turn off the signal converter before disconnecting the
cables. To reconnect the signal converter to the shared bus, connect the
cables before turning on the signal converter. Use the power switch to
turn off a standalone SCSI signal converter. To turn off an SBB SCSI
signal converter, pull it from its disk slot.
•
If you observe any “bus hung” messages, your DWZZA signal converters
may have the incorrect hardware. In addition, some DWZZA signal
converters that appear to have the correct hardware revision may cause
problems if they also have serial numbers in the range of CX444xxxxx
to CX449xxxxx.
To upgrade a DWZZA-AA or DWZZA-VA signal converter to the correct
revision, use the appropriate field change order (FCO), as follows:
–
–
DWZZA-AA-F002
DWZZA-VA-F001
2–8 Hardware Requirements and Restrictions
2.8 DS-DWZZH-03 and DS-DWZZH-05 UltraSCSI Hubs
The DS-DWZZH-03 and DS-DWZZH-05 series UltraSCSI hubs are the only
hubs supported in a TruCluster Server configuration. They are SCSI-2-
and draft SCSI-3-compliant SCSI 16-bit signal converters capable of data
transfer rates of up to 40 MB/sec.
These hubs can be listed with the other SCSI bus signal converters, but as
they are used differently in cluster configurations they will be discussed
differently in this manual.
A DS-DWZZH-03 or DS-DWZZH-05 UltraSCSI hub can be installed in:
•
A StorageWorks UltraSCSI BA356 shelf (which has the required
180-watt power supply).
•
The lower righthand device slot of the BA370 shelf within the RA7000
or ESA 10000 RAID array subsystems. This position minimizes cable
lengths and interference with disks.
•
A wide BA356 which has been upgraded to the 180-watt power supply
with the DS-BA35X-HH option.
A DS-DWZZH-03 or DS-DWZZH-05 UltraSCSI hub:
•
•
Improves the reliability of the detection of cable faults.
Provides for bus isolation of cluster systems while allowing the remaining
connections to continue to operate.
•
Allows for more separation of systems and storage in a cluster
configuration, because each SCSI bus segment can be up to 25 meters
in length. This allows a total separation of nearly 50 meters between
a system and the storage.
______________________
Note _______________________
The DS-DWZZH-03/05 UltraSCSI hubs cannot be connected to a
StorageWorks BA35X storage shelf because the storage shelf does
not provide termination power to the hub.
2.9 SCSI Cables
If you are using shared SCSI buses, you must determine if you need
cables with connectors that are low-density 50-pins, high-density 50-pins,
high-density 68-pins (HD68), or VHDCI (UltraSCSI). If you are using an
UltraSCSI hub, you will need HD68 to VHDCI and VHDCI to VHDCI cables.
In some cases, you also have the choice of straight or right-angle connectors.
Hardware Requirements and Restrictions 2–9
In addition, each supported cable comes in various lengths. Use the shortest
possible cables to adhere to the limits on SCSI bus length.
Table 2–3 describes each supported cable and the context in which you would
use the cable. Note that there are cables with the Compaq 6-3 part number
that are not listed, but are equivalent to the cables listed.
Table 2–3: Supported SCSI Cables
Pins
Cable
Connector
Density
Configuration Use
BN21W-0B
Three high
68-pin
A Y cable that can be attached to a
KZPSA-BB or KZPBA-CB if there
is no room for a trilink connector.
It can be used with a terminator to
provide external termination.
BN21M
One low, one
high
50-pin
LD to
68-pin
HD
Connects the single-ended end of
a DWZZA-AA or DWZZB-AA to
a TZ885 or TZ887.a
BN21K,
BN21L, or
328215-00X
Two HD68
68-pin
Connects BN21W Y cables or wide
devices. For example, connects
KZPBA-CBs, KZPSA-BBs, HSZ40s,
HSZ50s, the differential sides of
two SCSI signal converters, or a
DWZZB-AA to a BA356.
BN38C or
BN38D
One HD68, one HD68 to Connects a KZPBA-CB or KZPSA-BB
VHDCI
VHDCI
to a port on an UltraSCSI hub.
BN37A
Two VHDCI
VHDCI
to
VHDCI
Connects two VHDCI trilinks to
each other or an UltraSCSI hub to a
trilink on an HSZ70 or HSZ80
199629-002 or
189636-002
Two high
Two high
50-pin
HD to
68-pin
HD
Connect a Compaq 20/40 GB DLT
Tape Drive to a DWZZB-AA
146745-003 or
146776-003
50-pin
HD to
50-pin
HD
Daisy-chain two Compaq 20/40
GB DLT Tape Drives
a
Do not use a KZPBA-CB with a DWZZA-AA or DWZZB-AA and a TZ885 or TZ887. The DWZZAs and
DWZZBs can not operate at UltraSCSI speed.
Always check a SCSI cable for bent or broken pins. Be sure that you do not
bend or break any pins when you connect or disconnect a cable.
2–10 Hardware Requirements and Restrictions
2.10 SCSI Terminators and Trilink Connectors
Table 2–4 describes the supported trilink connectors and SCSI terminators
and the context in which you would use them.
Table 2–4: Supported SCSI Terminators and Trilink Connectors
Trilink
Connector or
Terminator
Den-
sity
Pins
Configuration Use
H885-AA
Three 68-pin Trilink connector that attaches to high-density,
68-pin cables or devices, such as a KZPSA-BB,
KZPBA-CB, HSZ40, HSZ50, or the differential
side of a SCSI signal converter. Can be
terminated with an H879-AA terminator to
provide external termination.
H8574-A or
H8860-AA
Low
50-pin Terminates a TZ885 or TZ887 tape drive.
341102-001
H879-AA
High
High
50-pin Terminates a Compaq 20/40 GB DLT Tape Drive
68-pin Terminates an H885-AA trilink connector
or BN21W-0B Y cable.
H8861-AA
VHDCI 68-pin VHDCI trilink connector that attaches to VHDCI
68-pin cables, UltraSCSI BA356 J A1, and HSZ70
or HSZ80 RAID controllers. Can be terminated
with an H8863-AA terminator if necessary.
H8863-AA
VHDCI 68-pin Terminate a VHDCI trilink connector.
The requirements for trilink connectors are as follows:
•
If you connect a SCSI cable to a trilink connector, do not block access to
the screws that mount the trilink, or you will be unable to disconnect the
trilink from the device without disconnecting the cable.
•
Do not install an H885-AA trilink if installing it will block an adjacent
peripheral component interconnect (PCI) port. Use a BN21W-0B Y cable
instead.
Hardware Requirements and Restrictions 2–11
3
Shared SCSI Bus Requirements and
Configurations Using UltraSCSI Hardware
A TruCluster Server cluster uses shared SCSI buses, external storage
shelves or RAID controllers, and supports disk mirroring and fast file system
recovery to provide high data availability and reliability.
This chapter:
•
•
•
Introduces SCSI bus configuration concepts
Describes requirements for the shared SCSI bus
Provides procedures for cabling TruCluster Server radial configurations
using UltraSCSI hubs and:
–
Dual-redundant HSZ70 or HSZ80 RAID array controllers enabled
for simultaneous failover
–
Dual-redundant HSZ70 or HSZ80 RAID array controllers enabled
for multiple-bus failover
•
Provides diagrams of TruCluster Server storage configurations using
UltraSCSI hardware configured for radial connections
______________________
Note _______________________
Although the UltraSCSI BA356 might have been included in
this chapter with the other UltraSCSI devices, it is not. The
UltraSCSI BA356 is covered in Chapter 9 with the configurations
using external termination. It cannot be cabled directly to an
UltraSCSI hub because it does not provide SCSI bus termination
power (termpwr).
In addition to using only supported hardware, adhering to the requirements
described in this chapter will ensure that your cluster operates correctly.
Chapter 9 contains additional information about using SCSI bus signal
converters, and also contains diagrams of TruCluster Server configurations
using UltraSCSI and non-UltraSCSI storage shelves and RAID array
controllers. The chapter also covers the older method of using external
Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware 3–1
termination and covers radial configurations with the DWZZH UltraSCSI
hubs and non-UltraSCSI RAID array controllers.
This chapter discusses the following topics:
•
•
•
•
•
•
•
Shared SCSI bus configuration requirements (Section 3.1)
SCSI bus performance (Section 3.2)
SCSI bus device identification numbers (Section 3.3)
SCSI bus length (Section 3.4)
SCSI bus termination (Section 3.5)
UltraSCSI hubs (Section 3.6)
Configuring UltraSCSI hubs with RAID array controllers (Section 3.7)
3.1 Shared SCSI Bus Configuration Requirements
A shared SCSI bus must adhere to the following requirements:
•
•
Only an external bus can be used for a shared SCSI bus.
SCSI bus specifications set a limit of 8 devices on an 8-bit (narrow) SCSI
bus. The limit is 16 devices on a 16-bit SCSI bus (wide). See Section 3.3
for more information.
•
•
The length of each physical bus is strictly limited. See Section 3.4 for
more information.
You can directly connect devices only if they have the same transmission
mode (differential or single-ended) and data path (narrow or wide). Use
a SCSI signal converter to connect devices with different transmission
modes. See Section 9.1 for information about the DWZZA (BA350) or
DWZZB (BA356) signal converters or the DS-BA35X-DA personality
module (which acts as a differential to single-ended signal converter
for the UltraSCSI BA356).
•
•
For each SCSI bus segment, you can have only two terminators, one
at each end. A physical SCSI bus may be composed of multiple SCSI
bus segments.
If you do not use an UltraSCSI hub, you must use trilink connectors
and Y cables to connect devices to a shared bus, so you can disconnect
the devices without affecting bus termination. See Section 9.2 for more
information.
•
Be careful when performing maintenance on any device that is on a
shared bus because of the constant activity on the bus. Usually, to
perform maintenance on a device without shutting down the cluster, you
must be able to isolate the device from the shared bus without affecting
bus termination.
3–2 Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware
•
All supported UltraSCSI host adapters support UltraSCSI disks at
UltraSCSI speeds in UltraSCSI BA356 shelves, RA7000 or ESA10000
storage arrays (HSZ70 and HSZ80), or RA8000 or ESA12000 storage
arrays (HSZ80 and HSG80). Older, non-UltraSCSI BA356 shelves are
supported with UltraSCSI host adapters and host RAID controllers as
long as they contain no UltraSCSI disks.
•
•
UltraSCSI drives and fast wide drives can be mixed together in an
UltraSCSI BA356 shelf (see Chapter 9).
Differential UltraSCSI adapters may be connected to either (or both)
a non-UltraSCSI BA356 shelf (via a DWZZB-VW) and the UltraSCSI
BA356 shelf (via the DS-BA35X-DA personality module) on the same
shared SCSI bus. The UltraSCSI adapter negotiates maximum transfer
speeds with each SCSI device (see Chapter 9).
•
The HSZ70 and HSZ80 UltraSCSI RAID controllers have a wide
differential UltraSCSI host bus with a Very High Density Cable
Interconnect (VHDCI) connector. HSZ70 and HSZ80 controllers will
work with fast and wide differential SCSI adapters (for example,
KZPSA-BB) at fast SCSI speeds.
•
•
Fast, wide SCSI drives (green StorageWorks building blocks (SBBs) with
part numbers ending in -VW) may be used in an UltraSCSI BA356 shelf.
Do not use fast, narrow SCSI drives (green SBBs with part numbers
ending in -VA) in any shelf that could assign the drive a SCSI ID greater
than 7. It will not work.
•
•
The UltraSCSI BA356 requires a 180-watt power supply (BA35X-HH).
It will not function properly with the older, lower-wattage BA35X-HF
universal 150-watt power supply (see Chapter 9).
An older BA356 that has been retrofitted with a BA35X-HH 180-watt
power supply and DS-BA35X-DA personality module is still only FCC
certified for Fast 10 configurations (see Chapter 9).
3.2 SCSI Bus Performance
Before you set up a SCSI bus, it is important that you understand a number
of issues that affect the viability of a bus and how the devices connected to it
operate. Specifically, bus performance is influenced by the following factors:
•
•
•
Transmission method
Data path
Bus speed
The following sections describe these factors.
Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware 3–3
3.2.1 SCSI Bus Versus SCSI Bus Segments
An UltraSCSI bus may be comprised of multiple UltraSCSI bus segments.
Each UltraSCSI bus segment is comprised of electrical conductors that
may be in a cable or a backplane, and cable or backplane connectors. Each
UltraSCSI bus segment must have a terminator at each end of the bus
segment.
Up to two UltraSCSI bus segments may be coupled together with UltraSCSI
hubs or signal converters, increasing the total length of the UltraSCSI bus.
3.2.2 Transmission Methods
Two transmission methods can be used in a SCSI bus:
•
Single-ended — In a single-ended SCSI bus, one data lead and one
ground lead are utilized for the data transmission. A single-ended
receiver looks only at the signal wire as the input. The transmitted
signal arrives at the receiving end of the bus on the signal wire
somewhat distorted by signal reflections. The length and loading of
the bus determine the magnitude of this distortion. This transmission
method is economical, but is more susceptible to noise than the
differential transmission method, and requires short cables. Devices
with single-ended SCSI devices include the following:
–
–
BA350, BA356, and UltraSCSI BA356 storage shelves
Single-ended side of a SCSI signal converter or personality module
•
Differential — Differential signal transmission uses two wires to
transmit a signal. The two wires are driven by a differential driver that
places a signal on one wire (+SIGNAL) and another signal that is 180
degrees out of phase (-SIGNAL) on the other wire. The differential
receiver generates a signal output only when the two inputs are different.
As signal reflections occur virtually the same on both wires, they are not
seen by the receiver, because it only sees differences on the two wires.
This transmission method is less susceptible to noise than single-ended
SCSI and enables you to use longer cables. Devices with differential
SCSI interfaces include the following:
–
–
–
–
KZPBA-CB
KZPSA-BB
HSZ40, HSZ50, HSZ70, and HSZ80 controllers
Differential side of a SCSI signal converter or personality module
You cannot use the two transmission methods in the same SCSI bus
segment. For example, a device with a differential SCSI interface must be
connected to another device with a differential SCSI interface. If you want to
3–4 Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware
connect devices that use different transmission methods, use a SCSI signal
converter between the devices. The DS-BA35X-DA personality module is
discussed in Section 9.1.2.2. See Section 9.1 for information about using the
DWZZ* series of SCSI signal converters.
You cannot use a DWZZA or DWZZB signal converter at UltraSCSI speeds
for TruCluster Server if there are any UltraSCSI disks on the bus, because
the DWZZA or DWZZB will not operate correctly at UltraSCSI speed.
The DS-BA35X-DA personality module contains a signal converter for
the UltraSCSI BA356. It is the interface between the shared differential
UltraSCSI bus and the UltraSCSI BA356 internal single-ended SCSI bus.
RAID array controller subsystems provide the function of a signal converter,
accepting the differential input and driving the single-ended device buses.
3.2.3 Data Path
There are two data paths for SCSI devices:
•
Narrow — Implies an 8-bit data path for SCSI-2. The performance of
this mode is limited.
•
Wide — Implies a 16-bit data path for SCSI-2 or UltraSCSI. This mode
increases the amount of data that is transferred in parallel on the bus.
3.2.4 Bus Speed
Bus speeds vary depending upon the bus clocking rate and bus width, as
shown in Table 3–1.
Table 3–1: SCSI Bus Speeds
Transfer Rate (MHz) Bus Width in Bytes Bus Bandwidth
SCSI Bus
(Speed) MB/sec
5
1
1
2
2
5
SCSI
10
10
20
10
20
40
Fast SCSI
Fast-Wide
UltraSCSI
3.3 SCSI Bus Device Identification Numbers
On a shared SCSI bus, each SCSI device uses a device address and must
have a unique SCSI ID (from 0 to 15). For example, each SCSI bus adapter
and each disk in a single-ended storage shelf uses a device address.
Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware 3–5
SCSI bus adapters have a default SCSI ID that you can change by using
console commands or utilities. For example, a KZPSA adapter has an initial
SCSI ID of 7.
______________________
Note _______________________
If you are using a DS-DWZZH-05 UltraSCSI hub with fair
arbitration enabled, SCSI ID numbering will change (see
Section 3.6.1.2).
Use the following priority order to assign SCSI IDs to the SCSI bus adapters
connected to a shared SCSI bus:
7-6-5-4-3-2-1-0-15-14-13-12-11-10-9-8
This order specifies that 7 is the highest priority, and 8 is the lowest priority.
When assigning SCSI IDs, use the highest priority ID for member systems
(starting at 7). Use lower priority IDs for disks.
Note that you will not follow this general rule when using the DS-DWZZH-05
UltraSCSI hub with fair arbitration enabled.
The SCSI ID for a disk in a BA350 storage shelf corresponds to its slot
location. The SCSI ID for a disk in a BA356 or UltraSCSI BA356 depends
upon its slot location and the personality module SCSI bus address switch
settings.
3.4 SCSI Bus Length
There is a limit to the length of the cables in a shared SCSI bus. The total
cable length for a SCSI bus segment is calculated from one terminated end
to the other.
If you are using devices that have the same transmission method and data
path (for example, wide differential), a shared bus will consist of only one bus
segment. If you have devices with different transmission methods, you will
have both single-ended and differential bus segments, each of which must be
terminated only at both ends and must adhere to the rules on bus length.
______________________
Note _______________________
In a TruCluster Server configuration, you always have
single-ended SCSI bus segments since all of the storage shelves
use a single-ended bus.
Table 3–2 describes the maximum cable length for a physical SCSI bus
segment.
3–6 Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware
Table 3–2: SCSI Bus Segment Length
SCSI Bus
Bus Speed
Maximum Cable Length
6 meters
Narrow, single-ended
Narrow, single-ended fast
Wide differential, fast
5 MB/sec
10 MB/sec
20 MB/sec
40 MB/sec
3 meters
25 meters
25 metersa
Differential UltraSCSI
a
The maximum separation between a host and the storage in a TruCluster Server configuration is 50
meters: 25 meters between any host and the UltraSCSI hub and 25 meters between the UltraSCSI hub
and the RAID array controller.
Because of the cable length limit, you must plan your hardware configuration
carefully, and ensure that each SCSI bus meets the cable limit guidelines.
In general, you must place systems and storage shelves as close together as
possible and choose the shortest possible cables for the shared bus.
3.5 Terminating the Shared SCSI Bus when Using UltraSCSI
Hubs
You must properly connect devices to a shared SCSI bus. In addition, you
can terminate only the beginning and end of each bus segment (either
single-ended or differential).
There are two rules for SCSI bus termination:
•
There are only two terminators for each SCSI bus segment. If you use an
UltraSCSI hub, you only have to install one terminator.
•
If you do not use an UltraSCSI hub, bus termination must be external.
External termination is covered in Section 9.2.
______________________ Notes ______________________
With the exception of the TZ885, TZ887, TL890, TL891, and
TL892, tape devices can only be installed at the end of a shared
SCSI bus. These tape devices are the only supported tape devices
that can be terminated externally.
We recommend that tape loaders be on a separate, shared SCSI
bus to allow normal shared SCSI bus termination for those shared
SCSI buses without tape loaders.
Whenever possible, connect devices to a shared bus so that they can be
isolated from the bus. This allows you to disconnect devices from the bus
for maintenance purposes, without affecting bus termination and cluster
operation. You also can set up a shared SCSI bus so that you can connect
additional devices at a later time without affecting bus termination.
Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware 3–7
Most devices have internal termination. For example, the UltraSCSI
KZPBA-CB and the fast and wide KZPSA-BB host bus adapters have
internal termination. When using a KZPBA-CB or KZPSA-BB with an
UltraSCSI hub, ensure that the onboard termination resistor SIPs have
not been removed.
You will need to provide termination at the storage end of one SCSI bus
segment. You will install an H8863-AA trilink connector on the HSZ70 or
HSZ80 at the bus end. Connect an H8861-AA terminator to the trilink
connector to terminate the bus.
Figure 3–1 shows a VHDCI trilink connector (UltraSCSI), which you may
attach to an HSZ70 or HSZ80.
Figure 3–1: VHDCI Trilink Connector (H8861-AA)
CXO5744A
3.6 UltraSCSI Hubs
The DS-DWZZH series UltraSCSI hubs are UltraSCSI signal converters
that provide radial connections of differential SCSI bus adapters and RAID
array controllers. Each connection forms a SCSI bus segment with SCSI bus
adapters or the storage unit. The hub provides termination for one end
of the bus segment. Termination for the other end of the bus segment is
provided by the:
•
•
Installed KZPBA-CB (or KZPSA-BB) termination resistor SIPs
External termination on a trilink connector attached to an UltraSCSI
BA356 personality module (DS-BA35X-DA), HSZ70, or HSZ80
______________________
Note _______________________
The DS-DWZZH-03/05 UltraSCSI hubs cannot be connected to a
StorageWorks BA35X storage shelf because the storage shelf does
not provide termination power to the hub.
3–8 Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware
3.6.1 Using a DWZZH UltraSCSI Hub in a Cluster Configuration
The DS-DWZZH-03 and DS-DWZZH-05 UltraSCSI hubs are supported in a
TruCluster Server cluster. They both provide radial connection of cluster
member systems and storage, and are similar in the following ways:
•
Contain internal termination for each port; therefore, the hub end of
each SCSI bus segment is terminated.
_____________________ Note _____________________
Do not put trilinks on a DWZZH UltraSCSI hub as it is not
possible to remove the DWZZH internal termination.
•
Require that termination power (termpwr) be provided by the SCSI bus
host adapters on each SCSI bus segment.
_____________________ Note _____________________
The UltraSCSI hubs are designed to sense loss of termination
power (such as a cable pull or termpwr not enabled on the
host adapter) and shut down the applicable port to prevent
corrupted signals on the remaining SCSI bus segments.
3.6.1.1 DS-DWZZH-03 Description
The DS-DWZZH-03:
•
•
Is a 3.5-inch StorageWorks building block (SBB)
Can be installed in:
–
A StorageWorks UltraSCSI BA356 storage shelf (which has the
required 180-watt power supply).
–
The lower righthand device slot of the BA370 shelf within the RA7000
or ESA 10000 RAID array subsystems. This position minimizes cable
lengths and interference with disks.
–
A non-UltraSCSI BA356 which has been upgraded to the 180-watt
power supply with the DS-BA35X-HH option.
•
Has three Very High Density Cable Interconnect (VHDCI) differential
SCSI bus connectors
•
•
Does not use a SCSI ID
Uses the storage shelf only to provide its power and mechanical support
(it is not connected to the shelf internal SCSI bus).
Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware 3–9
•
DS-DWZZH-03 and DS-DWZZH-05 UltraSCSI hubs may be housed
in the same storage shelf with disk drives. Table 3–3 provides the
supported configurations.
Figure 3–2 shows a front view of the DS-DWZZH-03 UltraSCSI hub.
Figure 3–2: DS-DWZZH-03 Front View
Differential symbol
ZK-1412U-AI
The differential symbol (and the lack of a single-ended symbol) indicates
that all three connectors are differential.
3.6.1.2 DS-DWZZH-05 Description
The DS-DWZZH-05:
•
•
Is a 5.25-inch StorageWorks building block (SBB)
Has five Very High Density Cable Interconnect (VHDCI) differential
SCSI bus connectors
•
Uses SCSI ID 7 whether or not fair arbitration mode is enabled.
Therefore, you cannot use SCSI ID 7 on the member systems’ SCSI bus
adapter.
The following section describes how to prepare the DS-DWZZH-05 UltraSCSI
hub for use on a shared SCSI bus in more detail.
3.6.1.2.1 DS-DWZZH-05 Configuration Guidelines
The DS-DWZZH-05 UltraSCSI hub can be installed in:
•
A StorageWorks UltraSCSI BA356 shelf (which has the required
180-watt power supply).
•
A non-UltraSCSI BA356 that has been upgraded to the 180-watt power
supply with the DS-BA35X-HH option.
3–10 Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware
_____________________ Note _____________________
Dual power supplies are recommended for any BA356 shelf
containing a DS-DWZZH-05 UltraSCSI hub in order to
provide a higher level of availability between cluster member
systems and storage.
•
The lower righthand device slot of the BA370 shelf within the RA7000
or ESA 10000 RAID array subsystems. This position minimizes cable
lengths and interference with disks.
A DS-DWZZH-05 UltraSCSI hub uses the storage shelf only to provide its
power and mechanical support (it is not connected to the shelf internal SCSI
bus).
______________________
Note _______________________
When the DS-DWZZH-05 is installed, its orientation is rotated
90 degrees counterclockwise from what is shown in Figure 3–3
and Figure 3–4.
The maximum configurations with combinations of DS-DWZZH-03 and
DS-DWZZH-05 UltraSCSI hubs, and disks in the same storage shelf
containing dual 180-watt power supplies, are shown in Table 3–3.
______________________
Note _______________________
With dual 180-watt power supplies installed, there are slots
available for six 3.5-inch SBBs or two 5.25-inch SBBs.
Table 3–3: DS-DWZZH UltraSCSI Hub Maximum Configurations
Disk Drivesa
Personality
Moduleb c
DS-DWZZH-03
DS-DWZZH-05
5
4
3
2
1
0
3
2
0
0
0
0
0
2
1
1
0
0
3
4
5
0
0
1
Not Installed
Installed
Installed
Installed
Installed
Not Installed
Not Installed
Installed
Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware 3–11
Table 3–3: DS-DWZZH UltraSCSI Hub Maximum Configurations (cont.)
Disk Drivesa
Personality
Moduleb c
DS-DWZZH-03
DS-DWZZH-05
1
1
1
2
3
Installed
Installed
0
a
DS-DWZZH UltraSCSI hubs and disk drives may coexist in a storage shelf. Installed disk drives are not
associated with the DS-DWZZH UltraSCSI hub SCSI bus segments; they are on the SCSI bus connected to
the personality module.
If the personality module is installed, you can install a maximum of four DS-DWZZH-03 UltraSCSI hubs.
The personality module must be installed to provide a path to any disks installed in the storage shelf.
b
c
3.6.1.2.2 DS-DWZZH-05 Fair Arbitration
Although each cluster member system and storage controller connected to an
UltraSCSI hub are on separate SCSI bus segments, they all share a common
SCSI bus and its bandwidth. As the number of systems accessing the storage
controllers increases, it is likely that the adapter with the highest priority
SCSI ID will obtain a higher proportion of the UltraSCSI bandwidth.
The DS-DWZZH-05 UltraSCSI hub provides a fair arbitration feature that
overrides the traditional SCSI bus priority. Fair arbitration applies only to
the member systems, not to the storage controllers (which are assigned
higher priority than the member system host adapters).
You enable fair arbitration by placing the switch on the front of the
DS-DWZZH-05 UltraSCSI hub to the Fair position (see Figure 3–4).
Fair arbitration works as follows. The DS-DWZZH-05 UltraSCSI hub is
assigned the highest SCSI ID, which is 7. During the SCSI arbitration phase,
the hub, because it has the highest priority, captures the SCSI ID of all host
adapters arbitrating for the bus. The hub compares the SCSI IDs of the host
adapters requesting use of the SCSI bus, and then allows the device with the
highest priority SCSI ID to take control of the SCSI bus. That SCSI ID is
removed from the group of captured SCSI IDs prior to the next comparison.
After the host adapter has been serviced, if there are still SCSI IDs retained
from the previous arbitration cycle, the next highest SCSI ID is serviced.
When all devices in the group have been serviced, the DS-DWZZH-05
repeats the sequence at the next arbitration cycle.
Fair arbitration is disabled by placing the switch on the front of the
DS-DWZZH-05 UltraSCSI hub in the Disable position (see Figure 3–4).
With fair arbitration disabled, the SCSI requests are serviced in the
conventional manner; the highest SCSI ID asserted during the arbitration
cycle obtains use of the SCSI bus.
3–12 Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware
______________________
Note _______________________
Host port SCSI ID assignments are not linked to the physical port
when fair arbitration is disabled.
The DS-DWZZH-05 reserves SCSI ID 7 regardless of whether fair
arbitration is enabled or not.
3.6.1.2.3 DS-DWZZH-05 Address Configurations
The DS-DWZZH-05 has two addressing modes: wide addressing mode and
narrow addressing mode. With either addressing mode, if fair arbitration is
enabled, each hub port is assigned a specific SCSI ID. This allows the fair
arbitration logic in the hub to identify the SCSI ID of the device participating
in the arbitration phase of the fair arbitration cycle.
_____________________ Caution
_____________________
If fair arbitration is enabled, The SCSI ID of the host adapter
must match the SCSI ID assigned to the hub port. Mismatching
or duplicating SCSI IDs will cause the hub to hang.
SCSI ID 7 is reserved for the DS-DWZZH-05 whether fair
arbitration is enabled or not.
J umper W1, accessible from the rear of the DS-DWZZH-05 (See Figure 3–3),
determines which addressing mode is used. The jumper is installed to select
narrow addressing mode. If fair arbitration is enabled, the SCSI IDs for the
host adapters are 0, 1, 2, and 3 (See the port numbers not in parentheses
in Figure 3–4). The controller ports are assigned SCSI IDs 4 through 6,
and the hub uses SCSI ID 7.
If jumper W1 is removed, the host adapter ports assume SCSI IDs 12,
13, 14, and 15. The controllers are assigned SCSI IDs 0 through 6. The
DS-DWZZH-05 retains the SCSI ID of 7.
Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware 3–13
Figure 3–3: DS-DWZZH-05 Rear View
W1
ZK-1448U-AI
3–14 Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware
Figure 3–4: DS-DWZZH-05 Front View
Fair Disable
Controller
Port
Host Port
SCSI ID
6 - 4
SCSI ID
2
(14)
(6 - 0)
Power
Host Port
SCSI ID
3
Busy
(15)
Host Port
SCSI ID
0
Host Port
SCSI ID
1
(12)
(13)
ZK-1447U-AI
3.6.1.2.4 SCSI Bus Termination Power
Each host adapter connected to a DS-DWZZH-05 UltraSCSI hub port must
supply termination power (termpwr) to enable the termination resistors
on each end of the SCSI bus segment. If the host adapter is disconnected
from the hub, the port is disabled. Only the UltraSCSI bus segment losing
termination power is affected. The remainder of the SCSI bus operates
normally.
3.6.1.2.5 DS-DWZZH-05 Indicators
The DS-DWZZH-05 has two indicators on the front panel (see Figure 3–4).
The green LED indicates that power is applied to the hub. The yellow LED
indicates that the SCSI bus is busy.
3.6.1.3 Installing the DS-DWZZH-05 UltraSCSI Hub
To install the DS-DWZZH-05 UltraSCSI hub, follow these steps:
1. Remove the W1 jumper to enable wide addressing mode (see Figure 3–3).
Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware 3–15
2. If fair arbitration is to be used, ensure that the switch on the front of
the DS-DWZZH-05 UltraSCSI hub is in the Fair position.
3. Install the DS-DWZZH-05 UltraSCSI hub in a UltraSCSI BA356,
non-UltraSCSI BA356 (if it has the required 180-watt power supply), or
BA370 storage shelf.
3.7 Preparing the UltraSCSI Storage Configuration
A TruCluster Server cluster provides you with high data availability through
the cluster file system (CFS), the device request dispatcher (DRD), service
failover through the cluster application availability (CAA) subsystem,
disk mirroring, and fast file system recovery. TruCluster Server supports
mirroring of the clusterwide root (/) file system, the member-specific boot
disks, and the cluster quorum disk through hardware RAID only. You can
mirror the clusterwide /usr and /var file systems and the data disks using
the Logical Storage Manager (LSM) technology. You must determine the
storage configuration that will meet your needs. Mirroring disks across two
shared buses provides the most highly available data.
See the TruCluster Server Software Product Description (SPD) to determine
the supported storage shelves, disk devices, and RAID array controllers.
Disk devices used on the shared bus must be installed in a supported storage
shelf or behind a RAID array controller. Before you connect a storage
shelf to a shared SCSI bus, you must install the disks in the unit. Before
connecting a RAID array controller to a shared SCSI bus, install the disks
and configure the storagesets. For detailed information about installation
and configuration, see your storage shelf (or RAID array controller)
documentation.
______________________
Note _______________________
The following sections mention only the KZPBA-CB UltraSCSI
host bus adapter because it is needed to obtain UltraSCSI speeds
for UltraSCSI configurations. The KZPSA-BB host bus adapter
may be used in any configuration in place of the KZPBA-CB
without any cable changes. Be aware though, the KZPSA-BB is
not an UltraSCSI device and therefore only works at fast-wide
speed (20 MB/sec).
The following sections describe how to prepare and install cables for storage
configurations on a shared SCSI bus using UltraSCSI hubs and the HSZ70
or HSZ80 RAID array controllers.
3–16 Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware
3.7.1 Configuring Radially Connected TruCluster Server Clusters
with UltraSCSI Hardware
Radial configurations with RAID array controllers allow you to take
advantage of the benefits of hardware mirroring, and to achieve a
no-single-point-of-failure (NSPOF) cluster. Typical RAID array storage
subsystems used in TruCluster Server cluster configurations are:
•
•
•
RA7000 or ESA10000 with HSZ70 controller
RA7000 or ESA10000 with HSZ80 controller
RA8000 or ESA12000 with HSZ80 controller
When used with TruCluster Server, one advantage of using a RAID array
controller is the ability to hardware mirror the clusterwide root (/) file
system, member system boot disks, swap disk, and quorum disk. When used
in a dual-redundant configuration, Tru64 UNIX Version 5.0A supports both
transparent failover, which occurs automatically, without host intervention,
and multiple-bus failover, which requires host intervention for some failures.
______________________
Note _______________________
Enable mirrored cache for dual-redundant configurations to
further ensure the availability of unwritten cache data.
Use transparent failover if you only have one shared SCSI bus. Both
controllers are connected to the same host and device buses, and either
controller can service all of the units if the other controller fails.
Transparent failover compensates only for a controller failure, and not
for failures of either the SCSI bus or host adapters and is therefore not a
NSPOF configuration.
______________________
Note _______________________
Set each controller to transparent failover mode before configuring
devices (SET FAILOVER COPY = THIS_CONTROLLER).
To achieve a NSPOF configuration, you need multiple-bus failover and two
shared SCSI buses.
You may use multiple-bus failover (SET MULTIBUS_FAILOVER COPY =
THIS_CONTROLLER) to help achieve a NSPOF configuration if each host has
two shared SCSI buses to the array controllers. One SCSI bus is connected
to one controller and the other SCSI bus is connected to the other controller.
Each member system has a host bus adapter for each shared SCSI bus. The
load can be distributed across the two controllers. In case of a host adapter
Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware 3–17
or SCSI bus failure, the host can redistribute the load to the surviving
controller. In case of a controller failure, the surviving controller will handle
all units.
______________________ Notes ______________________
Multiple-bus failover does not support device partitioning with
the HSZ70 or HSZ80.
Partioned storagesets and partitioned single-disk units cannot
function in multiple-bus failover dual-redundant configurations.
Because they are not supported, you must delete your partitions
before configuring the HSZ70 or HSZ80 controllers for
multiple-bus failover.
Device partitioning is supported with HSG80 array controllers
with ACS Version 8.5.
Multiple-bus failover does not support tape drives or CD-ROM
drives.
The following sections describe how to cable the HSZ70 or HSZ80 for
TruCluster Server configurations. See Chapter 6 for information regarding
Fibre Channel storage.
3.7.1.1 Preparing an HSZ70 or HSZ80 for a Shared SCSI Bus Using Transparent
Failover Mode
When using transparent failover mode:
•
•
Both controllers of an HSZ70 are connected to the same shared SCSI bus
For an HSZ80:
–
–
–
Port 1 of controller A and Port 1 of controller B are on the same
SCSI bus.
If used, Port 2 of controller A and Port 2 of controller B are on the
same SCSI bus.
HSZ80 targets assigned to Port 1 cannot be seen by Port 2.
To cable a dual-redundant HSZ70 or HSZ80 for transparent failover in a
TruCluster Server configuration using a DS-DWZZH-03 or DS-DWZZH-05
UltraSCSI hub, see Figure 3–5 (HSZ70) or Figure 3–6 (HSZ80) and follow
these steps:
1. You will need two H8861-AA VHDCI trilink connectors. Install an
H8863-AA VHDCI terminator on one of the trilinks.
3–18 Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware
2. Attach the trilink with the terminator to the controller that you want
to be on the end of the shared SCSI bus. Attach an H8861-AA VHDCI
trilink connector to:
•
•
HSZ70 controller A and controller B
HSZ80 Port 1 (2) of controller A and Port 1 (2) of controller B
___________________ Note
___________________
You must use the same port on each HSZ80 controller.
3. Install a BN37A cable between the trilinks on:
•
•
HSZ70 controller A and controller B
HSZ80 controller A Port 1 (2) and controller B Port 1 (2)
The BN37A-0C is a 0.3-meter cable and the BN37A-0E is a 0.5-meter
cable.
4. Install the DS-DWZZH-03 or DS-DWZZH-05 UltraSCSI hub in
an UltraSCSI BA356, non-UltraSCSI BA356 (with the required
180-watt power supply), or BA370 storage shelf (see Section 3.6.1.1
or Section 3.6.1.2).
5. If you are using a:
•
DWZZH-03: Install a BN37A cable between any DWZZH-03 port
and the open trilink connector on HSZ70 controller A (B) or HSZ80
controller A Port 1 (2) or controller B Port 1 (2).
•
DWZZH-05:
–
–
–
Verify that the fair arbitration switch is in the Fair position to
enable fair arbitration (see Section 3.6.1.2.2).
Ensure that the W1 jumper is removed to select wide addressing
mode (see Section 3.6.1.2.3)
Install a BN37A cable between the DWZZH-05 controller port
and the open trilink connector on HSZ70 controller A (B) or
HSZ80 controller A Port 1 (2) or controller B Port 1 (2).
6. When the KZPBA-CB host bus adapters in each member system are
installed, connect each KZPBA-CB to a DWZZH port with a BN38C (or
BN38D) HD68 to VHDCI cable. Ensure that the KZPBA-CB SCSI ID
matches the SCSI ID assigned to the DWZZH-05 port it is cabled to
(12, 13, 14, and 15).
Figure 3–5 shows a two-member TruCluster Server configuration with a
radially connected dual-redundant HSZ70 RAID array controller configured
for transparent failover.
Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware 3–19
Figure 3–5: Shared SCSI Bus with HSZ70 Configured for Transparent
Failover
Network
Member
System
1
Member
System
2
Memory
Channel
Interface
Memory Channel
KZPBA-CB (ID 6)
Memory Channel
KZPBA-CB (ID 7)
T
T
1
1
T
DS-DWZZH-03
T
T
3
2
2
4
3
T
Controller A
HSZ70
Controller B
HSZ70
StorageWorks
RAID Array 7000
ZK-1599U-AI
Table 3–4 shows the components used to create the clusters shown in
Figure 3–5, Figure 3–6, Figure 3–7, and Figure 3–8.
3–20 Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware
Table 3–4: Hardware Components Used in Configuration Shown in Figure
3–5 Through Figure 3–8
Description
Callout Number
BN38C cablea
1
2
3
BN37A cableb
H8861-AA VHDCI trilink connector
H8863-AA VHDCI terminatorb
4
a
b
The maximum length of the BN38C (or BN38D) cable on one SCSI bus segment must not exceed 25 meters.
The maximum combined length of the BN37A cables must not exceed 25 meters.
Figure 3–6 shows a two-member TruCluster Server configuration with a
radially connected dual-redundant HSZ80 RAID array controller configured
for transparent failover.
Figure 3–6: Shared SCSI Bus with HSZ80 Configured for Transparent
Failover
Network
Member
System
1
Member
System
2
Memory
Channel
Interface
Memory Channel
KZPBA-CB (ID 6)
Memory Channel
KZPBA-CB (ID 7)
T
T
1
1
T
DS-DWZZH-03
T
T
2
4
3
2
3
T
Port 1
Port 2
Port 2
Port 1
Controller B
HSZ80
Controller A
HSZ80
StorageWorks
RAID Array 8000
ZK-1600U-AI
Table 3–4 shows the components used to create the cluster shown in
Figure 3–6.
Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware 3–21
3.7.1.2 Preparing a Dual-Redundant HSZ70 or HSZ80 for a Shared SCSI Bus Using
Multiple-Bus Failover
Multiple-bus failover is a dual-redundant controller configuration in which
each host has two paths (two shared SCSI buses) to the array controller
subsystem. The host(s) have the capability to move LUNs from one controller
(shared SCSI bus) to the other. If one host adapter or SCSI bus fails, the
host(s) can move all storage to the other path. Because both controllers
can service all of the units, either controller can continue to service all of
the units if the other controller fails. Therefore, multiple-bus failover can
compensate for a failed host bus adapter, SCSI bus, or RAID array controller,
and can, if the rest of the cluster has necessary hardware, provide a NSPOF
configuration.
______________________
Note _______________________
Each host (cluster member system) requires at least two
KZPBA-CB host bus adapters.
Although both the HSZ70 and HSZ80 have multiple-bus failover, it operates
differently:
•
HSZ70: Only one controller (or shared SCSI bus) is active for the
units that are preferred (assigned) to it. If all units are preferred to
one controller, then all units are accessed through one controller. If a
controller detects a problem, all of its units are failed over to the other
controller. If the host detects a problem with the host bus adapter or SCSI
bus, the host initiates the failover to the other controller (and SCSI bus).
•
HSZ80: Both HSZ80 controllers can be active at the same time. If the
host detects a problem with a host bus adapter or SCSI bus, the host
initiates the failover to the other controller. If a controller detects a
problem, all of its units are failed over to the other controller.
Also, the HSZ80 has two ports on each controller. If multiple-bus failover
mode is enabled, the targets assigned to any one port are visible to
all ports unless access to a unit is restricted to a particular port (on a
unit-by-unit basis).
To cable an HSZ70 or HSZ80 for multiple-bus failover in a TruCluster Server
configuration using DS-DWZZH-03 or DS-DWZZH-05 UltraSCSI hubs (you
need two hubs), see Figure 3–7 (HSZ70) and Figure 3–8 (HSZ80) and follow
these steps:
1. Install an H8863-AA VHDCI terminator on each of two H8861-AA
VHDCI trilink connectors.
2. Install H8861-AA VHDCI trilink connectors (with terminators) on:
3–22 Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware
•
•
HSZ70 controller A and controller B
HSZ80 controller A Port 1 (2) and controller B Port 1 (2)
___________________ Note
___________________
You must use the same port on each HSZ80 controller.
3. Install the DS-DWZZH-03 or DS-DWZZH-05 UltraSCSI hub in a
DS-BA356, BA356 (with the required 180-watt power supply), or BA370
storage shelf (see Section 3.6.1.1 or Section 3.6.1.2)
4. If you are using a:
•
DS-DWZZH-03: Install a BN37A VHDCI to VHDCI cable between
the trilink connector on controller A (HSZ70) or controller A Port 1
(2) (HSZ80) and any DS-DWZZH-03 port. Install a second BN37A
cable between the trilink on controller B (HSZ70) or controller B
Port 1 (2) (HSZ80) and any port on the second DS-DWZZH-03.
•
DS-DWZZH-05:
–
–
–
Verify that the fair arbitration switch is in the Fair position to
enable fair arbitration (see Section 3.6.1.2.2)
Ensure that the W1 jumper is removed to select wide addressing
mode (see Section 3.6.1.2.3)
Install a BN37A cable between the DWZZH-05 controller port
and the open trilink connector on HSZ70 controller A or HSZ80
controller A Port 1 (2)
–
Install a second BN37A cable between the second DWZZH-05
controller port and the open trilink connector on HSZ70
controller B or HSZ80 controller B Port 1 (2)
5. When the KZPBA-CBs are installed, use a BN38C (or BN38D) HD68
to VHDCI cable to connect the first KZPBA-CB on each system to a
port on the first DWZZH hub. Ensure that the KZPBA-CB SCSI ID
matches the SCSI ID assigned to the DWZZH-05 port it is cabled to
(12, 13, 14, and 15).
6. Install BN38C (or BN38D) HD68 to VHDCI cables to connect the second
KZPBA-CB on each system to a port on the second DWZZH hub. Ensure
that the KZPBA-CB SCSI ID matches the SCSI ID assigned to the
DWZZH-05 port it is cabled to (12, 13, 14, and 15).
Figure 3–7 shows a two-member TruCluster Server configuration with
a radially connected dual-redundant HSZ70 configured for multiple-bus
failover.
Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware 3–23
Figure 3–7: TruCluster Server Configuration with HSZ70 in Multiple-Bus
Failover Mode
Network
Member
System
1
Member
System
2
Memory
Channel
Interface
Memory Channel
KZPBA-CB (ID 6)
KZPBA-CB (ID 6)
Memory Channel
T
KZPBA-CB (ID 7)
KZPBA-CB (ID 7)
T
T
T
1
1
1
1
T
T
DS-DWZZH-03
DS-DWZZH-03
T
T
T
T
3
2
3
2
4
T
4
T
Controller A
HSZ70
Controller B
HSZ70
StorageWorks
RAID Array 7000
ZK-1601U-AI
Table 3–4 shows the components used to create the cluster shown in
Figure 3–7.
Figure 3–8 shows a two-member TruCluster Server configuration with
a radially connected dual-redundant HSZ80 configured for multiple-bus
failover.
3–24 Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware
Figure 3–8: TruCluster Server Configuration with HSZ80 in Multiple-Bus
Failover Mode
Networks
Memory
Channel
Interfaces
Member System 2
Member System 1
Memory Channel (mca1)
Memory Channel (mca0)
Memory Channel (mca1)
Memory Channel (mca0)
KZPBA-CB (ID 6)
KZPBA-CB (ID 6)
T
KZPBA-CB (ID 7)
KZPBA-CB (ID 7)
T
T
T
1
1
1
1
T
T
DS-DWZZH-03
DS-DWZZH-03
T
T
2
T
T
2
3
4
3
4
T
T
Port 1
Port 2
Port 2
Port 1
Controller B
HSZ80
Controller A
HSZ80
StorageWorks
RAID Array 8000
ZK-1602U-AI
Table 3–4 shows the components used to create the cluster shown in
Figure 3–8.
Shared SCSI Bus Requirements and Configurations Using UltraSCSI Hardware 3–25
4
TruCluster Server System Configuration
Using UltraSCSI Hardware
This chapter describes how to prepare systems for a TruCluster Server
cluster, using UltraSCSI hardware and the preferred method of radial
configuration, including how to connect devices to a shared SCSI bus for
the TruCluster Server product. This chapter does not provide detailed
information about installing devices; it describes only how to set up the
hardware in the context of the TruCluster Server product. Therefore, you
must have the documentation that describes how to install the individual
pieces of hardware. This documentation should arrive with the hardware.
All systems in the cluster must be connected via the Memory Channel cluster
interconnect. Not all members must be connected to a shared SCSI bus.
You need to allocate disks for the following uses:
•
One or more disks to hold the Tru64 UNIX operating system. The disk(s)
are either private disk(s) on the system that will become the first cluster
member, or disk(s) on a shared bus that the system can access.
•
•
•
One or more disks on a shared SCSI bus to hold the clusterwide root (/),
/usr, and /var AdvFS file systems.
One disk per member, normally on a shared SCSI bus, to hold member
boot partitions.
Optionally, one disk on a shared SCSI bus to act as the quorum disk. See
Section 1.4.1.4, and for a more detailed discussion of the quorum disk,
see the TruCluster Server Cluster Administration manual.
All configurations covered in this manual assume the use of a shared SCSI
bus.
______________________
Note _______________________
If you are using Fibre Channel storage, see Chapter 6.
Before you connect devices to a shared SCSI bus, you must:
•
Plan your hardware configuration, determining which devices will be
connected to each shared SCSI bus, which devices will be connected
together, and which devices will be at the ends of each bus.
TruCluster Server System Configuration Using UltraSCSI Hardware 4–1
This is especially critical if you will install tape devices on the shared
SCSI bus. With the exception of the TZ885, TZ887, TL890, TL891, and
TL892, tape devices can only be installed at the end of a shared SCSI
bus. These tape devices are the only supported tape devices that can
be terminated externally.
•
•
Place the devices as close together as possible and ensure that shared
SCSI buses will be within length limitations.
Prepare the systems and storage shelves for the appropriate bus
connection, including installing SCSI controllers, UltraSCSI hubs,
trilink connectors, and SCSI signal converters.
After you install all necessary cluster hardware and connect the shared
SCSI buses, be sure that the systems can recognize and access all the shared
disks (see Section 4.3.2). You can then install the TruCluster Server software
as described in the TruCluster Server Software Installation manual.
4.1 Planning Your TruCluster Server Hardware
Configuration
Before you set up a TruCluster Server hardware configuration, you must
plan a configuration to meet your performance and availability needs. You
must determine the following components for your configuration:
•
Number and type of member systems and the number of shared SCSI
buses
You can use two to eight member systems for TruCluster Server. A
greater number of member systems connected to shared SCSI buses
gives you better application performance and more availability. However,
all the systems compete for the same buses to service I/O requests, so a
greater number of systems decreases I/O performance.
Each member system must have a supported SCSI adapter for each
shared SCSI bus connection. There must be enough PCI slots for the
Memory Channel cluster interconnect(s) and SCSI adapters. The number
of available PCI slots depends on the type of AlphaServer system.
•
Cluster interconnects
You need only one cluster interconnect in a cluster. For TruCluster
Server Version 5.0A, the cluster interconnect is the Memory Channel.
However, you can use redundant cluster interconnects to protect against
an interconnect failure and for easier hardware maintenance. If you
have more than two member systems, you must have one Memory
Channel hub for each interconnect.
4–2 TruCluster Server System Configuration Using UltraSCSI Hardware
•
Number of shared SCSI buses and the storage on each shared bus
Using shared SCSI buses increases storage availability. You can connect
32 shared SCSI buses to a cluster member. You can use any combination
of KZPSA-BB, KZPBA-CB, or KGPSA-BC/CA host bus adapters.
In addition, RAID array controllers allow you to increase your storage
capacity and protect against disk, controller, host bus adapter, and SCSI
bus failures. Mirroring data across shared buses provides you with more
reliable and available data. You can use Logical Storage Manager (LSM)
host-based mirroring for all storage except the clusterwide root (/) file
system, the member-specific boot disks, and the swap and quorum disk.
•
•
No single-point-of-failure (NSPOF) TruCluster Server cluster
You can use mirroring and multiple-bus failover with the HSZ70, HSZ80,
and HSG80 RAID array controllers to create a NSPOF TruCluster Server
cluster (providing the rest of the hardware is installed).
Tape loaders on a shared SCSI bus
Because of the length of the internal SCSI cables in some tape
loaders (up to 3 meters), they cannot be externally terminated with a
trilink/terminator combination. Therefore, in general, with the exception
of the TL890, TL891, and TL892, tape loaders must be on the end of the
shared SCSI bus. See Chapter 8 for information on configuring tape
devices on a shared SCSI bus.
•
You cannot use Prestoserve in a TruCluster Server cluster to cache I/O
operations for any storage device, regardless of whether it is located
on a shared bus or a bus local to a given system. Because data in
the Prestoserve buffer cache of one member is not accessible to other
member systems, TruCluster Server cannot provide correct failover when
Prestoserve is being used.
Table 4–1 describes how to maximize performance, availability, and
storage capacity in your TruCluster Server hardware configuration. For
example, if you want greater application performance without decreasing
I/O performance, you can increase the number of member systems or you
can set up additional shared storage.
Table 4–1: Planning Your Configuration
To increase:
You can:
Application performance
I/O performance
Increase the number of member systems.
Increase the number of shared buses.
Increase the number of member systems.
Member system availability
Cluster interconnect availability Use redundant cluster interconnects.
TruCluster Server System Configuration Using UltraSCSI Hardware 4–3
Table 4–1: Planning Your Configuration (cont.)
To increase:
You can:
Disk availability
Mirror disks across shared buses.
Use a RAID array controller.
Increase the number of shared buses.
Use a RAID array controller.
Increase disk size.
Shared storage capacity
4.2 Obtaining the Firmware Release Notes
You may be required to update the system or SCSI controller firmware
during a TruCluster Server installation, so you may need the firmware
release notes.
You can obtain the firmware release notes from:
•
The Web at the following URL:
http://www.compaq.com/support/
Select Alpha Systems from the downloadable drivers &
utilities menu. Then select the appropriate system.
•
The current Alpha Systems Firmware Update CD-ROM.
_____________________ Note _____________________
To obtain the firmware release notes from the Firmware
Update Utility CD-ROM, your kernel must be configured for
the ISO 9660 Compact Disk File System (CDFS).
To obtain the release notes for the firmware update, follow these steps:
1. At the console prompt, or using the system startup log if the Tru64
UNIX operating system is running, determine the drive number of
the CD-ROM.
2. Boot the Tru64 UNIX operating system if it is not already running.
3. Log in as root.
4. Place the Alpha Systems Firmware Update CD-ROM applicable to
the Tru64 UNIX version installed (or to be installed) into the drive.
5. Mount the CD-ROM as follows (/dev/disk/cdrom0c is used as an
example CD-ROM drive):
# mount -rt cdfs -o noversion /dev/disk/cdrom0cc /mnt
4–4 TruCluster Server System Configuration Using UltraSCSI Hardware
6. Copy the appropriate release notes to your system disk. In this
example, obtain the firmware release notes for the AlphaServer
DS20 from the Version 5.6 Alpha Firmware Update CD-ROM:
# cp /mnt/doc/ds20_v56_fw_relnote.txt ds20-rel-notes
7. Unmount the CD-ROM drive:
# umount /mnt
8. Print the release notes.
4.3 TruCluster Server Hardware Installation
Member systems may be connected to a shared SCSI bus with a peripheral
component interconnect (PCI) SCSI adapter. Before you install a PCI SCSI
adapter into a PCI slot on a member system, ensure that the module is at
the correct hardware revision.
The qualification and use of the DS-DWZZH-series UltraSCSI hubs in
TruCluster Server clusters allows the PCI host bus adapters to be cabled
into a cluster in two different ways:
•
Preferred method with radial connection to a DWZZH UltraSCSI hub
and internal termination: The PCI host bus adapter internal termination
resistor SIPs are not removed. The host bus adapters and storage
subsystems are connected directly to a DWZZH UltraSCSI hub port.
There can be only one member system connected to a hub port.
The use of a DWZZH UltraSCSI hub in a TruCluster Server cluster is
preferred because it improves the reliability to detect cable faults.
•
Old method with external termination: Shared SCSI bus termination is
external to the PCI host adapters. This is the old method used to connect
a PCI host adapter to the cluster; remove the adapter termination
resistor SIPs and install a Y cable and an H879-AA terminator for
external termination. This allows the removal of a SCSI bus cable from
the host adapter without affecting SCSI bus termination.
This method (discussd in Chapter 9 and Chapter 10) may be used with or
without a DWZZH UltraSCSI hub. When used with an UltraSCSI hub,
there may be more than one member system on a SCSI bus segment
attached to a DS-DWZZH-03 hub port.
The following sections describe how to install the KZPBA-CB
PCI-to-UltraSCSI differential host adapter and configure them into
TruCluster Server clusters using the preferred method of radial connection
with internal termination.
TruCluster Server System Configuration Using UltraSCSI Hardware 4–5
______________________
Note _______________________
The KZPSA-BB can be used in any configuration in place of the
KZPBA-CB. The use of the KZPSA-BB is not mentioned in this
chapter because it is not UltraSCSI hardware, and it cannot
operate at UltraSCSI speeds.
The use of the KZPSA-BB (and the KZPBA-CB) with external termination is
covered in Chapter 10.
It is assumed that when you start to install the hardware necessary to create
a TruCluster Server configuration, you have sufficient storage to install the
TruCluster Server software, and that you have set up any RAID storagesets.
Follow the steps in Table 4–2 to start the procedure for TruCluster Server
hardware installation. You can save time by installing the Memory Channel
adapters, redundant network adapters (if applicable), and KZPBA-CB SCSI
adapters all at the same time.
Follow the directions in the referenced documentation, or the steps in
the referenced tables, returning to the appropriate table when you have
completed the steps in the referenced table.
_____________________ Caution
_____________________
Static electricity can damage modules and electronic components.
We recommend using a grounded antistatic wrist strap and a
grounded work surface when handling modules.
Table 4–2: Configuring TruCluster Server Hardware
Action
Refer to:
Step
Chapter 5a
1
Install the Memory Channel module(s),
cables, and hub(s) (if a hub is required).
2
Install Ethernet or FDDI network
adapters.
User ’s guide for the applicable
Ethernet or FDDI adapter,
and the user ’s guide for the
applicable system
Install ATM adapters if using ATM.
Chapter 7 and ATMworks 350
Adapter Installation and Service
3
Install a KZPBA-CB UltraSCSI adapter
for each radially connected shared SCSI
bus in each member system.
Section 4.3.1 and Table 4–3
4–6 TruCluster Server System Configuration Using UltraSCSI Hardware
Table 4–2: Configuring TruCluster Server Hardware (cont.)
Action
Refer to:
Step
4
Update the system SRM console
Use the firmware update release
firmware from the latest Alpha Systems notes (Section 4.2)
Firmware Update CD-ROM.
______________________ Note
_____________________
The SRM console firmware includes the ISP1020/1040-based PCI
option firmware, which includes the KZPBA-CB. When you update the
SRM console firmware, you are enabling the KZPBA-CB firmware to
be updated. On a powerup reset, the SRM console loads KZPBA-CB
adapter firmware from the console system flash ROM into NVRAM for
all Qlogic ISP1020/1040-based PCI options, including the KZPBA-CB
PCI-to-Ultra SCSI adapter.
a
If you install additional KZPBA-CB SCSI adapters or an extra network adapter at this time, delay testing
the Memory Channel until you have installed all of the hardware.
4.3.1 Installation of a KZPBA-CB Using Internal Termination for a
Radial Configuration
Use this method of cabling member systems and shared storage in a
TruCluster Server cluster if you are using a DWZZH UltraSCSI hub. You
must reserve at least one hub port for shared storage.
The DWZZH-series UltraSCSI hubs are designed to allow more separation
between member systems and shared storage. Using the UltraSCSI hub also
improves the reliability of the detection of cable faults.
A side benefit is the ability to connect the member systems’ SCSI adapter
directly to a hub port without external termination. This simplifies the
configuration by reducing the number of cable connections.
A DWZZH UltraSCSI hub can be installed in:
•
A StorageWorks UltraSCSI BA356 shelf that has the required 180-watt
power supply.
•
The lower righthand device slot of the BA370 shelf within the RA7000
or ESA 10000 RAID array subsystems. This position minimizes cable
lengths and interference with disks.
•
A non-UltraSCSI BA356 that has been upgraded to the 180-watt power
supply with the DS-BA35X-HH option.
An UltraSCSI hub only receives power and mechanical support from the
storage shelf. There is no SCSI bus continuity between the DWZZH and
storage shelf.
TruCluster Server System Configuration Using UltraSCSI Hardware 4–7
The DWZZH contains a differential to single-ended signal converter for each
hub port (sometimes referred to as a DWZZA on a chip, or DOC chip). The
single-ended sides are connected together to form an internal single-ended
SCSI bus segment. Each differential SCSI bus port is terminated internal to
the DWZZH with terminators that cannot be disabled or removed.
Power for the DWZZH termination (termpwr) is supplied by the host SCSI
bus adapter or RAID array controller connected to the DWZZH port. If the
member system or RAID array controller is powered down, or the cable is
removed from the KZPBA-CB, RAID array controller, or hub port, the loss of
termpwr disables the hub port without affecting the remaining hub ports
or SCSI bus segments. This is similar to removing a Y cable when using
external termination.
______________________
Note _______________________
The UltraSCSI BA356 DS-BA35X-DA personality module does not
generate termpwr. Therefore, you cannot connect an UltraSCSI
BA356 directly to a DWZZH hub. The use of the UltraSCSI
BA356 in a TruCluster Server cluster is discussed in Chapter 9.
The other end of the SCSI bus segment is terminated by the KZPBA-CB
onboard termination resistor SIPs, or by a trilink connector/terminator
combination installed on the RAID array controller.
The KZPBA-CB UltraSCSI host adapter:
•
Is a high-performance PCI option connecting the PCI-based host system
to the devices on a 16-bit, ultrawide differential SCSI bus.
•
•
•
Is installed in a PCI slot of the supported member system.
Is a single-channel, ultrawide differential adapter.
Operates at the following speeds:
–
–
–
–
5 MB/sec narrow SCSI at slow speed
10 MB/sec narrow SCSI at fast speed
20 MB/sec wide differential SCSI
40 MB/sec wide differential UltraSCSI
______________________
Note _______________________
Even though the KZPBA-CB is an UltraSCSI device, it has an
HD68 connector.
4–8 TruCluster Server System Configuration Using UltraSCSI Hardware
Your storage shelves or RAID array subsystems should be set up before
completing this portion of an installation.
Use the steps in Table 4–3 to set up a KZPBA-CB for a TruCluster Server
cluster that uses radial connection to a DWZZH UltraSCSI hub.
Table 4–3: Installing the KZPBA-CB for Radial Connection to a DWZZH
UltraSCSI Hub
Action
Refer to:
Step
1
Ensure that the eight KZPBA-CB internal
Section 4.3.1, Figure 4–1,
termination resistor SIPs, RM1-RM8 are installed. and KZPBA-CB
PCI-to-Ultra SCSI
Differential Host Adapter
User ’s Guide
2
3
Power down the system. Install a KZPBA-CB
PCI-to-UltraSCSI differential host adapter in
the PCI slot corresponding to the logical bus to
be used for the shared SCSI bus. Ensure that
the number of adapters are within limits for the
system, and that the placement is acceptable.
TruCluster Server
Cluster Administration,
Section 2.4.2, and
KZPBA-CB PCI-to-Ultra
SCSI Differential Host
Adapter User ’s Guide
Install a BN38C cable between the KZPBA-CB
UltraSCSI host adapter and a DWZZH port.
_____________________
Notes
_____________________
The maximum length of a SCSI bus segment is 25 meters, including the
bus length internal to the adapter and storage devices.
One end of the BN38C cable is 68-pin high density. The other end is
68-pin VHDCI. The DWZZH accepts the 68-pin VHDCI connector.
The number of member systems in the cluster has to be one less than
the number of DWZZH ports.
4
5
Power up the system and use the show config
and show device console commands to display
the installed devices and information about
the KZPBA-CBs on the AlphaServer systems.
Look for QLogic ISP1020 in the show config
display and isp in the show device display to
determine which devices are KZPBA-CBs.
Section 4.3.2 and
Example 4–1 through
Example 4–4
Use the show pk* or show isp* console
commands to determine the KZPBA-CB SCSI
bus ID, and then use the set console command
to set the SCSI bus ID.
Section 4.3.3 and
Example 4–5 through
Example 4–7
TruCluster Server System Configuration Using UltraSCSI Hardware 4–9
Table 4–3: Installing the KZPBA-CB for Radial Connection to a DWZZH
UltraSCSI Hub (cont.)
Action
Refer to:
Step
_____________________
Notes
_____________________
Ensure that the SCSI ID that you use is distinct from all other SCSI IDs
on the same shared SCSI bus. If you do not remember the other SCSI
IDs, or do not have them recorded, you must determine these SCSI IDs.
If you are using a DS-DWZZH-05, you cannot use SCSI ID 7
for a KZPBA-CB UltraSCSI adapter; SCSI ID 7 is reserved for
DS-DWZZH-05 use.
If you are using a DS-DWZZH-05 and fair arbitration is enabled, you
must use the SCSI ID assigned to the hub port the adapter is connected
to.
You will have problems if you have two or more SCSI adapters at the
same SCSI ID on any one SCSI bus.
6
7
Repeat steps 1 through 6 for any other
KZPBA-CBs to be installed on this shared SCSI
bus on other member systems.
Connect a DS-DWZZH-03 or DS-DWZZH-05
UltraSCSI hub to an:
Section 3.6
HSZ70 or HSZ80 in transparent failover mode
HSZ70 or HSZ80 in multiple-bus failover mode
Section 3.7.1.1
Section 3.7.1.2
4.3.2 Displaying KZPBA-CB Adapters with the show Console
Commands
Use the show config and show device console commands to display
system configuration. Use the output to determine which devices are
KZPBA-CBs, and to determine their SCSI bus IDs.
Example 4–1 shows the output from the show config console command on
an AlphaServer DS20 system.
Example 4–1: Displaying Configuration on an AlphaServer DS20
P00>>> show config
AlphaServer DS20 500 MHz
SRM Console:
PALcode:
T5.4-15
OpenVMS PALcode V1.54-43, Tru64 UNIX PALcode V1.49-45
Processors
CPU 0
Alpha 21264-4 500 MHz
Bcache size: 4 MB
SROM Revision: V1.82
SROM Revision: V1.82
CPU 1
Alpha 21264-4 500 MHz
Bcache size: 4 MB
4–10 TruCluster Server System Configuration Using UltraSCSI Hardware
Example 4–1: Displaying Configuration on an AlphaServer DS20 (cont.)
Core Logic
Cchip
Dchip
Pchip 0
Pchip 1
DECchip 21272-CA Rev 2.1
DECchip 21272-DA Rev 2.0
DECchip 21272-EA Rev 2.2
DECchip 21272-EA Rev 2.2
TIG
Rev 4.14
Arbiter
Rev 2.10 (0x1)
MEMORY
Array #
-------
0
Size
Base Addr
---------- ---------
512 MB
000000000
Total Bad Pages = 0
Total Good Memory = 512 MBytes
PCI Hose 00
Bus 00 Slot 05/0: Cypress 82C693
Bridge to Bus 1, ISA
Bus 00 Slot 05/1: Cypress 82C693 IDE
dqa.0.0.105.0
Bus 00 Slot 05/2: Cypress 82C693 IDE
dqb.0.1.205.0
Bus 00 Slot 05/3: Cypress 82C693 USB
Bus 00 Slot 07: DECchip 21152-AA
Bus 00 Slot 08: QLogic ISP1020
Bridge to Bus 2, PCI
pkc0.7.0.8.0
dkc0.0.0.8.0
SCSI Bus ID 7
HSZ70
HSZ70
HSZ70
HSZ70CCL
HSZ70
HSZ70
HSZ70
HSZ70
HSZ70
HSZ70
dkc1.0.0.8.0
dkc100.1.0.8.0
dkc101.1.0.8.0
dkc2.0.0.8.0
dkc3.0.0.8.0
dkc4.0.0.8.0
dkc5.0.0.8.0
dkc6.0.0.8.0
dkc7.0.0.8.0
Bus 00 Slot 09: QLogic ISP1020
pkd0.7.0.9.0
dkd0.0.0.9.0
dkd1.0.0.9.0
dkd100.1.0.9.0
dkd101.1.0.9.0
dkd102.1.0.9.0
SCSI Bus ID
HSZ40
HSZ40
HSZ40
HSZ40
7
HSZ40
.
.
.
dkd5.0.0.9.0
dkd6.0.0.9.0
dkd7.0.0.9.0
HSZ40
HSZ40
HSZ40
Bus 02 Slot 00: NCR 53C875
pka0.7.0.2000.0
dka0.0.0.2000.0
dka100.1.0.2000.0
dka200.2.0.2000.0
dka500.5.0.2000.0
SCSI Bus ID
RZ1CB-CS
RZ1CB-CS
RZ1CB-CS
RRD46
7
7
Bus 02 Slot 01: NCR 53C875
pkb0.7.0.2001.0
SCSI Bus ID
TruCluster Server System Configuration Using UltraSCSI Hardware 4–11
Example 4–1: Displaying Configuration on an AlphaServer DS20 (cont.)
Bus 02 Slot 02: DE500-AA Network Controller
ewa0.0.0.2002.0
00-06-2B-00-0A-48
PCI Hose 01
Bus 00 Slot 07: DEC PCI FDDI
fwa0.0.0.7.1
08-00-2B-B9-0D-5D
Rev: 22, mca0
Bus 00 Slot 08: DEC PCI MC
Bus 00 Slot 09: DEC PCI MC
Rev: 22, mcb0
ISA
Slot
0
Device Name
Type
Enabled BaseAddr IRQ
DMA
0
1
2
3
4
5
MOUSE
KBD
COM1
COM2
LPT1
FLOPPY
Embedded
Embedded
Embedded
Embedded
Embedded
Embedded
Yes
Yes
Yes
Yes
Yes
Yes
60
60
3f8
2f8
3bc
3f0
12
1
4
3
7
6
2
Example 4–2 shows the output from the show device console command
entered on an AlphaServer DS20 system.
Example 4–2: Displaying Devices on an AlphaServer DS20
P00>>> show device
dka0.0.0.2000.0
dka100.1.0.2000.0
dka200.2.0.2000.0
dka500.5.0.2000.0
dkc0.0.0.8.0
DKA0
RZ1CB-CS 0656
RZ1CB-CS 0656
RZ1CB-CS 0656
RRD46 1337
DKA100
DKA200
DKA500
DKC0
HSZ70 V71Z
dkc1.0.0.8.0
DKC1
HSZ70 V71Z
.
.
.
dkc7.0.0.8.0
dkd0.0.0.9.0
dkd1.0.0.9.0
dkd100.1.0.9.0
dkd101.1.0.9.0
dkd102.1.0.9.0
DKC7
DKD0
DKD1
DKD100
DKD101
DKD102
HSZ70 V71Z
HSZ40 YA03
HSZ40 YA03
HSZ40 YA03
HSZ40 YA03
HSZ40 YA03
.
.
.
dkd7.0.0.9.0
dva0.0.0.0.0
ewa0.0.0.2002.0
fwa0.0.0.7.1
pka0.7.0.2000.0
pkb0.7.0.2001.0
pkc0.7.0.8.0
pkd0.7.0.9.0
DKD7
DVA0
EWA0
FWA0
PKA0
PKB0
PKC0
PKD0
HSZ40 YA03
00-06-2B-00-0A-48
08-00-2B-B9-0D-5D
SCSI Bus ID 7
SCSI Bus ID 7
SCSI Bus ID 7 5.57
SCSI Bus ID 7 5.57
4–12 TruCluster Server System Configuration Using UltraSCSI Hardware
Example 4–3 shows the output from the show config console command
entered on an AlphaServer 8200 system.
Example 4–3: Displaying Configuration on an AlphaServer 8200
>>> show config
Name
Type
Rev
Mnemonic
TLSB
4++
5+
KN7CC-AB
MS7CC
KFTIA
8014
5000
2020
0000
0000
0000
kn7cc-ab0
ms7cc0
kftia0
8+
C0 Internal PCI connected to kftia0
pci0
isp0
isp1
0+ QLogic ISP1020 10201077
1+ QLogic ISP1020 10201077
2+ DECchip 21040-AA 21011
4+ QLogic ISP1020 10201077
5+ QLogic ISP1020 10201077
6+ DECchip 21040-AA 21011
0001
0001
0023 tulip0
0001
0001
0023 tulip1
isp2
isp3
C1 PCI connected to kftia0
0+ KZPAA
11000
0001 kzpaa0
1+ QLogic ISP1020 10201077
0005
isp4
2+ KZPSA
3+ KZPSA
4+ KZPSA
7+ DEC PCI MC
81011
81011
81011
181011
0000 kzpsa0
0000 kzpsa1
0000 kzpsa2
000B
mc0
Example 4–4 shows the output from the show device console command
entered on an AlphaServer 8200 system.
Example 4–4: Displaying Devices on an AlphaServer 8200
>>> show device
polling for units on isp0, slot0, bus0, hose0...
polling for units on isp1, slot1, bus0, hose0...
polling for units on isp2, slot4, bus0, hose0...
polling for units on isp3, slot5, bus0, hose0...
polling for units kzpaa0, slot0, bus0, hose1...
pke0.7.0.0.1
dke0.0.0.0.1
dke200.2.0.0.1
dke400.4.0.0.1
kzpaa4
DKE0
DKE200
DKE400
SCSI Bus ID 7
RZ28
442D
442D
0064
RZ28
RRD43
polling for units isp4, slot1, bus0, hose1...
dkf0.0.0.1.1
dkf1.0.0.1.1
dkf2.0.0.1.1
dkf3.0.0.1.1
DKF0
DKF1
DKF2
DKF3
HSZ70
HSZ70
HSZ70
HSZ70
V70Z
V70Z
V70Z
V70Z
TruCluster Server System Configuration Using UltraSCSI Hardware 4–13
Example 4–4: Displaying Devices on an AlphaServer 8200 (cont.)
dkf4.0.0.1.1
dkf5.0.0.1.1
dkf6.0.0.1.1
dkf100.1.0.1.1
dkf200.2.0.1.1
dkf300.3.0.1.1
DKF4
DKF5
DKF6
DKF100
DKF200
DKF300
HSZ70
HSZ70
HSZ70
RZ28M
RZ28M
RZ28
V70Z
V70Z
V70Z
0568
0568
442D
polling for units on kzpsa0, slot 2, bus 0, hose1...
kzpsa0.4.0.2.1
dkg0.0.0.2.1
dkg1.0.0.2.1
dkg2.0.0.2.1
dkg100.1.0.2.1
dkg200.2.0.2.1
dkg300.3.0.2.1
dkg
DKG0
DKG1
DKG2
DKG100
DKG200
DKG300
TPwr 1 Fast 1 Bus ID 7
HSZ50-AX X29Z
L01 A11
HSZ50-AX X29Z
HSZ50-AX X29Z
RZ26N
RZ28
RZ26N
0568
392A
0568
polling for units on kzpsa1, slot 3, bus 0, hose1...
kzpsa1.4.0.3.1
dkh100.1.0.3.1
dkh200.2.0.3.1
dkh300.3.0.3.1
dkh
TPwr 1 Fast 1 Bus ID 7
L01 A11
DKH100
DKH200
DKH300
RZ28
RZ26
442D
392A
442D
RZ26L
polling for units on kzpsa2, slot 4, bus 0, hose1...
kzpsa2.4.0.4.1
dki100.1.0.3.1
dki200.2.0.3.1
dki300.3.0.3.1
dki
TPwr 1 Fast 1 Bus ID 7
L01 A10
DKI100
DKI200
DKI300
RZ26
RZ28
RZ26
392A
442C
392A
4.3.3 Displaying Console Environment Variables and Setting the
KZPBA-CB SCSI ID
The following sections show how to use the show console command to display
the pk* and isp* console environment variables, and set the KZPBA-CB
SCSI ID on various AlphaServer systems. Use these examples as guides
for your system.
Note that the console environment variables used for the SCSI options vary
from system to system. Also, a class of environment variables (for example,
pk* or isp*) may show both internal and external options.
Compare the following examples with the devices shown in the show
config and show dev examples to determine which devices are KZPSA-BBs
or KZPBA-CBs on the shared SCSI bus.
4–14 TruCluster Server System Configuration Using UltraSCSI Hardware
4.3.3.1 Displaying KZPBA-CB pk* or isp* Console Environment Variables
To determine the console environment variables to use, execute the show
pk* and show isp* console commands.
Example 4–5 shows the pk console environment variables for an AlphaServer
DS20.
Example 4–5: Displaying the pk* Console Environment Variables on an
AlphaServer DS20 System
P00>>>show pk*
pka0_disconnect
pka0_fast
pka0_host_id
1
1
7
pkb0_disconnect
pkb0_fast
pkb0_host_id
1
1
7
pkc0_host_id
7
pkc0_soft_term
on
pkd0_host_id
7
pkd0_soft_term
on
Comparing the show pk* command display in Example 4–5 with the
show config command in Example 4–1, you determine that the first two
devices shown in Example 4–5, pkao and pkb0 are for NCR 53C875 SCSI
controllers. The next two devices, pkc0 and pkd0, shown in Example 4–1 as
Qlogic ISP1020 devices, are KZPBA-CBs, which are really Qlogic ISP1040
devices (regardless of what the console says).
Our interest then, is in pkc0 and pkd0.
Example 4–5 shows two pk*0_soft_term environment variables,
pkc0_soft_term and pkd0_soft_term, both of which are on.
The pk*0_soft_term environment variable applies to systems using the
QLogic ISP1020 SCSI controller, which implements the 16-bit wide SCSI
bus and uses dynamic termination.
The QLogic ISP1020 module has two terminators, one for the 8 low bits and
one for the high 8 bits. There are five possible values for pk*0_soft_term:
•
•
•
off — Turns off both low 8 bits and high 8 bits
low — Turns on low 8 bits and turns off high 8 bits
high — Turns on high 8 bits and turns off low 8 bits
TruCluster Server System Configuration Using UltraSCSI Hardware 4–15
•
•
on — Turns on both low 8 bits and high 8 bits
diff — Places the bus in differential mode
The KZPBA-CB is a Qlogic ISP1040 module, and its termination is
determined by the presence or absence of internal termination resistor SIPs
RM1-RM8. Therefore, the pk*0_soft_term environment variable has no
meaning and it may be ignored.
Example 4–6 shows the use of the show isp console command to display
the console environment variables for KZPBA-CBs on an AlphaServer 8x00.
Example 4–6: Displaying Console Variables for a KZPBA-CB on an
AlphaServer 8x00 System
P00>>>show isp*
isp0_host_id
7
isp0_soft_term
on
isp1_host_id
7
isp1_soft_term
on
isp2_host_id
7
isp2_soft_term
on
isp3_host_id
7
isp3_soft_term
on
isp5_host_id
7
isp5_soft_term
diff
Both Example 4–3 and Example 4–4 show five isp devices; isp0, isp1,
isp2, isp3, and isp4. In Example 4–6, the show isp* console command
shows isp0, isp1, isp2, isp3, and isp5.
The console code that assigns console environment variables counts every I/O
adapter including the KZPAA, which is the device after isp3, and therefore
logically isp4 in the numbering scheme. The show isp console command
skips over isp4 because the KZPAA is not a QLogic 1020/1040 class module.
Example 4–3 and Example 4–4 show that isp0, isp1, isp2, and isp3
are devices on the internal KFTIA PCI bus and not on a shared SCSI bus.
Only isp4, the KZPBA-CB, is on a shared SCSI bus (and the show isp
console command displays it as isp5). The other three shared SCSI buses
use KZPSA-BBs (Use the show pk* console command to display the KZPSA
console environment variables.)
4–16 TruCluster Server System Configuration Using UltraSCSI Hardware
4.3.3.2 Setting the KZPBA-CB SCSI ID
After you determine the console environment variables for the KZPBA-CBs
on the shared SCSI bus, use the set console command to set the SCSI
ID. For a TruCluster Server cluster, you will most likely have to set the
SCSI ID for all KZPBA-CB UltraSCSI adapters except one. And, if you are
using a DS-DWZZH-05, you will have to set the SCSI IDs for all KZPBA-CB
UltraSCSI adapters.
______________________ Notes ______________________
You will have problems if you have two or more SCSI adapters at
the same SCSI ID on any one SCSI bus.
If you are using a DS-DWZZH-05, you cannot use SCSI ID 7
for a KZPBA-CB UltraSCSI adapter; SCSI ID 7 is reserved for
DS-DWZZH-05 use.
If DS-DWZZH-05 fair arbitration is enabled, The SCSI ID of the
host adapter must match the SCSI ID assigned to the hub port.
Mismatching or duplicating SCSI IDs will cause the hub to hang.
SCSI ID 7 is reserved for the DS-DWZZH-05 whether fair arbitration is
enabled or not.
Use the set console command as shown in Example 4–7 to set the SCSI ID.
In this example, the SCSI ID is set for KZPBA-CB pkc on the AlphaServer
DS20 shown in Example 4–5.
Example 4–7: Setting the KZPBA-CB SCSI Bus ID
P00>>> show pkc0_host_id
7
P00>>> set pkc0_host_id 6
P00>>> show pkc0_host_id
6
4.3.3.3 KZPBA-CB Termination Resistors
The KZPBA-CB internal termination is disabled by removing the
termination resistors RM1-RM8, as shown in Figure 4–1.
TruCluster Server System Configuration Using UltraSCSI Hardware 4–17
Figure 4–1: KZPBA-CB Termination Resistors
Internal Narrow Device
Connector P2
Internal Wide Device
Connector J2
JA1
SCSI Bus Termination
Resistors RM1-RM8
ZK-1451U-AI
4–18 TruCluster Server System Configuration Using UltraSCSI Hardware
5
Setting Up the Memory Channel Cluster
Interconnect
This chapter describes Memory Channel configuration restrictions, and
describes how to set up the Memory Channel cluster interconnect, including
setting up a Memory Channel hub, Memory Channel optical converter (MC2
only), and connecting link cables.
Two versions of the Memory Channel PCI adapter are available; CCMAA
and CCMAB (MC2).
Two variations of the CCMAA PCI adapter are in use; CCMAA-AA (MC1)
and CCMAA-AB (MC1.5). As the hardware used with these two PCI
adapters is the same, this manual often refers to MC1 when referring to
either of these variations.
See the TruCluster Server Software Product Description (SPD) for a list
of the supported Memory Channel hardware. See the Memory Channel
User ’s Guide for illustrations and more detailed information about installing
jumpers, Memory Channel adapters, and hubs.
You can have two Memory Channel adapters with TruCluster Server, but
only one rail can be active at a time. This is referred to as a failover pair. If
the active rail fails, cluster communications fails over to the inactive rail.
See Section 2.2 for a discussion on Memory Channel restrictions.
To set up the Memory Channel interconnects, follow these steps, referring to
the appropriate section and the Memory Channel User ’s Guide as necessary:
1. Set the Memory Channel jumpers (Section 5.1).
2. Install the Memory Channel adapter into a PCI slot on each system
(Section 5.2).
3. If you are using fiber optics with MC2, install the CCMFB fiber optics
module (Section 5.3).
4. If you have more than two systems in the cluster, install a Memory
Channel hub (Section 5.4).
5. Connect the Memory Channel cables (Section 5.5).
6. After you complete steps 1 through 5 for all systems in the cluster, apply
power to the systems and run Memory Channel diagnostics (Section 5.6).
Setting Up the Memory Channel Cluster Interconnect 5–1
____________________
Note _____________________
If you are installing SCSI or network adapters, you may wish
to complete all hardware installation before powering up the
systems to run Memory Channel diagnostics.
5.1 Setting the Memory Channel Adapter Jumpers
The meaning of the Memory Channel adapter module jumpers depends upon
the version of the Memory Channel module.
5.1.1 MC1 and MC1.5 Jumpers
The MC1 and MC1.5 modules (CCMAA-AA and CCMAA-AB respectively)
have adapter jumpers that designate whether the configuration is using
standard or virtual hub mode. If virtual hub mode is being used, there can
be only two systems. One system must be virtual hub 0 (VH0) and the other
must be virtual hub 1 (VH1).
The Memory Channel adapter should arrive with the jumpers set for
standard hub mode (pins 1 to 2 jumpered). Confirm that the jumpers are
set properly for your configuration. The jumper configurations are shown
as if you were holding the module with the jumpers facing you, with the
module end plate in your left hand. The jumpers are right next to the
factory/maintenance cable connector, and are described in Table 5–1.
Table 5–1: MC1 and MC1.5 Jumper Configuration
If hub mode is:
Jumper:
Example:
Standard
Pins 1 to 2
12 3
Virtual: VH0
Virtual: VH1
Pins 2 to 3
12 3
12 3
None needed; store the jumper on pin 1 or 3
5–2 Setting Up the Memory Channel Cluster Interconnect
If you are upgrading from virtual hub mode to standard hub mode (or from
standard hub mode to virtual hub mode), be sure to change the jumpers on
all Memory Channel adapters on the rail.
5.1.2 MC2 Jumpers
The MC2 module (CCMAB) has multiple jumpers. They are numbered right
to left, starting with J 1 in the upper righthand corner (as you view the
jumper side of the module with the endplate in your left hand). The leftmost
jumpers are J 11 and J 10. J 11 is above J 10.
Most of the jumper settings are straightforward, but the window size
jumper, J 3, needs some explanation.
If a CCMAA adapter (MC1 or MC1.5) is installed, 128 MB of address space is
allocated for Memory Channel use. If a CCMAB adapter (MC2) PCI adapter
is installed, the memory space allocation for Memory Channel depends on
the J 3 jumper and can be 128 or 512 MB.
If two Memory Channel adapters are used as a failover pair to provide
redundancy, the address space allocated for the logical rail depends on the
smaller window size of the physical adapters.
During a rolling upgrade from an MC1 failover pair to an MC2 failover pair,
the MC2 modules can be jumpered for 128 MB or 512 MB. If jumpered
for 512 MB, the increased address space is not achieved until all MC PCI
adapters have been upgraded and the use of 512 MB is enabled. On one
member system, use the sysconfig command to reconfigure the Memory
Channel kernel subsystem to initiate the use of 512 MB address space. The
configuration change is propagated to the other cluster member systems
by entering the following command:
# /sbin/sysconfig -r rm rm_use_512=1
See the TruCluster Server Cluster Administration manual for more
information on failover pairs.
The MC2 jumpers are described in Table 5–2.
Table 5–2: MC2 Jumper Configuration
Jumper:
Description:
Example:
J 1: Hub Mode
Standard: Pins 1 to 2
12 3
Setting Up the Memory Channel Cluster Interconnect 5–3
Table 5–2: MC2 Jumper Configuration (cont.)
Jumper:
Description:
Example:
VH0: Pins 2 to 3
12 3
VH1: None needed;
store the jumper on
pin 1 or pin 3
12 3
12 3
12 3
12 3
12 3
12 3
12 3
J 3: Window Size
512 MB: Pins 2 to 3
128 MB: Pins 1 to 2
J 4: Page Size
8-KB page size (UNIX):
Pins 1 to 2
4-KB page size (not
used): Pins 2 to 3
J 5: AlphaServer
8x00 Mode
8x00 mode selected:
Pins 1 to 2a
8x00 mode not selected:
Pins 2 to 3
5–4 Setting Up the Memory Channel Cluster Interconnect
Table 5–2: MC2 Jumper Configuration (cont.)
Jumper:
Description:
Example:
J 10 and J 11: Fiber Fiber Off: Pins 1 to 2
Optics Mode Enable
3
2
1
Fiber On: Pins 2 to
3 pins
3
2
1
a
Increases the maximum sustainable bandwidth for 8x00 systems. If the jumpers are in this position for
other systems, the bandwidth is decreased.
The MC2 linecard (CCMLB) has two jumpers, J 2 and J 3, that are used
to enable fiber optics mode. The jumpers are located near the middle of
the module (as you view the jumper side of the module with the endplate
in your left hand). J umper J 2 is on the right. The MC2 linecard jumpers
are described in Table 5–3.
Table 5–3: MC2 Linecard Jumper Configurations
Jumper:
Description:
Example:
J 2 and J 3: Fiber
Mode
Fiber Off: Pins 2 to 3
12 3
Fiber On: Pins 1 to 2
12 3
5.2 Installing the Memory Channel Adapter
Install the Memory Channel adapter in an appropriate peripheral component
interconnect (PCI) slot (see Section 2.2). Secure the module at the backplane.
Ensure that the screw is tight to maintain proper grounding.
The Memory Channel adapter comes with a straight extension plate. This
fits most systems; however, you may have to replace the extender with an
angled extender (AlphaServer 2100A, for instance), or for an AlphaServer
8200/8400, GS60, GS60E, or GS140 remove the extender completely.
Setting Up the Memory Channel Cluster Interconnect 5–5
If you are setting up a redundant Memory Channel configuration, install the
second Memory Channel adapter right after installing the first Memory
Channel adapter. Ensure that the jumpers are correct and are the same
on both modules.
After you install the Memory Channel adapter(s), replace the system panels
unless you have more hardware to install.
5.3 Installing the MC2 Optical Converter in the Member
System
If you are going to use a CCMFB optical converter along with the MC2
PCI adapter, install it at the same time that you install the MC2 CCMAB.
To install a MC2 CCMFB optical converter in the member system, follow
these steps. See Section 5.5.2.4 if you are installing an optical converter
in an MC2 hub.
1. Remove the bulkhead blanking plate for the desired PCI slot.
2. Thread one end of the fiber optics cable (BN34R) through the PCI
bulkhead slot.
3. Thread the optics cable through the slot in the optical converter module
(CCMFB) endplate (at the top of the endplate).
4. Remove the cable tip protectors and attach the keyed plug to the
connector on the optical converter module. Tie-wrap the cable to the
module.
5. Seat the optical converter module firmly into the PCI backplane and
secure the module with the PCI card cage mounting screw.
6. Attach the 1-meter BN39B-01 cable from the CCMAB Memory Channel
2 PCI adapter to the CCMFB optical converter.
7. Route the fiber optics cable to the remote system or hub.
8. Repeat steps 1 through 7 for the optical converter on the second system.
See Section 5.5.2.4 if you are installing an optical converter in an MC2
hub.
5.4 Installing the Memory Channel Hub
You may use a hub in a two-node TruCluster Server cluster, but the hub is
not required. When there are more than two systems in a cluster, you must
use a Memory Channel hub as follows:
•
For use with the MC1 or MC1.5 CCMAA adapter, you must install the
hub within 10 meters of each of the systems.
5–6 Setting Up the Memory Channel Cluster Interconnect
For use with the MC2 CCMAB adapter, the hub must be placed within
4 or 10 meters (the length of the BN39B link cables) of each system. If
fiber optics is used in conjunction with the MC2 adapter, the hub may be
placed up to 31 meters from the systems.
•
•
Ensure that the voltage selection switch on the back of the hub is set to
select the correct voltage for your location (115V or 230V).
Ensure that the hub contains a linecard for each system in the cluster
(the hub comes with four linecards) as follows:
–
–
CCMLA linecards for the CCMHA MC1 hub
CCMLB linecards for the CCMHB MC2 hub. Note that the linecards
cannot be installed in the opto only slot.
•
•
•
•
If you have a four-node cluster, you may want to install an extra linecard
for troubleshooting use.
If you have an eight-node cluster, all linecards must be installed in the
same hub.
For MC2, if fiber optics converters are used, they can only be installed in
hub slots opto only, 0/opto, 1/opto, 2/opto, and 3/opto.
If you have a five-node or greater MC2 cluster using fiber optics, you
will need two or three CCMHB hubs, depending on the number of fiber
optics connections. You will need one hub for the CCMLB linecards (and
possible optics converters) and up to two hubs for the CCMFB optics
converter modules. The CCMHB-BA hub has no linecards.
5.5 Installing the Memory Channel Cables
Memory Channel cable installation depends on the Memory Channel module
revision, and whether or not you are using fiber optics. The following sections
describe how to install the Memory Channel cables for MC1 and MC2.
5.5.1 Installing the MC1 or MC1.5 Cables
To set up an MC1 or MC1.5 interconnect, use the BC12N-10 10-meter
link cables to connect Memory Channel adapters and, optionally, Memory
Channel hubs.
______________________
Note _______________________
Do not connect an MC1 or MC1.5 link cable to an MC2 module.
Setting Up the Memory Channel Cluster Interconnect 5–7
5.5.1.1 Connecting MC1 or MC1.5 Link Cables in Virtual Hub Mode
For an MC1 virtual hub configuration (two nodes in the cluster), connect the
BC12N-10 link cables between the Memory Channel adapters installed in
each of the systems.
_____________________ Caution
_____________________
Be very careful when installing the link cables. Insert the cables
straight in.
Gently push the cable’s connector into the receptacle, and then use the
screws to pull the connector in tight. The connector must be tight to ensure
a good ground contact.
If you are setting up redundant interconnects, all Memory Channel adapters
in a system must have the same jumper setting, either VH0 or VH1.
______________________
Note _______________________
With the TruCluster Server Version 5.0A product and virtual hub
mode, there is no longer a restriction requiring that mca0 in one
system be connected to mca0 in the other system.
5.5.1.2 Connecting MC1 Link Cables in Standard Hub Mode
If there are more than two systems in a cluster, use a standard hub
configuration. Connect a BC12N-10 link cable between the Memory Channel
adapter and a linecard in the CCMHA hub, starting at the lowest numbered
slot in the hub.
If you are setting up redundant interconnects, the following restrictions
apply:
•
•
Each adapter installed in a system must be connected to a different hub.
Each Memory Channel adapter in a system must be connected to
linecards that are installed in the same slot position in each hub. For
example, if you connect one adapter to a linecard installed in slot 1 in
one hub, you must connect the other adapter in that system to a linecard
installed in slot 1 of the second hub.
Figure 5–1 shows Memory Channel adapters connected to linecards that are
in the same slot position in the Memory Channel hubs.
5–8 Setting Up the Memory Channel Cluster Interconnect
Figure 5–1: Connecting Memory Channel Adapters to Hubs
Memory Channel
hub 1
System A
Memory Channel
hub 2
Linecards
Memory
Channel
adapters
ZK-1197U-AI
5.5.2 Installing the MC2 Cables
To set up an MC2 interconnect, use the BN39B-04 (4-meter) or BN39B-10
(10-meter) link cables for virtual hub or standard hub configurations without
optical converters.
If optical converters are used, use the BN39B-01 1-meter link cable and the
BN34R-10 (10-meter) or BN34R-31 (31-meter) fiber optics cable.
5.5.2.1 Installing the MC2 Cables for Virtual Hub Mode Without Optical Converters
To set up an MC2 configuration for virtual hub mode, use BN39B-04
(4-meter) or BN39B-10 (10-meter) Memory Channel link cables to connect
Memory Channel adapters to each other.
______________________ Notes ______________________
MC2 link cables (BN39B) are black cables.
Do not connect an MC2 cable to an MC1 or MC1.5 CCMAA
module.
Setting Up the Memory Channel Cluster Interconnect 5–9
Gently push the cable’s connector into the receptacle, and then use the
screws to pull the connector in tight. The connector must be tight to ensure
a good ground contact.
If you are setting up redundant interconnects, all Memory Channel adapters
in a system must have the same jumper setting, either VH0 or VH1.
5.5.2.2 Installing MC2 Cables in Virtual Hub Mode Using Optical Converters
If you are using optical converters in an MC2 configuration, install an optical
converter module (CCMFB) when you install the CCMAB Memory Channel
PCI adapter in each system in the virtual hub configuration. You should
also connect the CCMAB Memory Channel adapter to the optical converter
with a BN39B-01 cable. When you install the CCMFB optical converter
module in the second system, you connect the two systems with the BN34R
fiber optics cable (see Section 5.3).
5.5.2.3 Connecting MC2 Link Cables in Standard Hub Mode (No Fiber Optics)
If there are more than two systems in a cluster, use a Memory Channel
standard hub configuration. Connect a BN39B-04 (4-meter) or BN39B-10
(10-meter) link cable between the Memory Channel adapter and a linecard
in the CCMHB hub, starting at the lowest numbered slot in the hub.
If you are setting up redundant interconnects, the following restrictions
apply:
•
•
Each adapter installed in a system must be connected to a different hub.
Each Memory Channel adapter in a system must be connected to
linecards that are installed in the same slot position in each hub. For
example, if you connect one adapter to a linecard installed in slot 0/opto
in one hub, you must connect the other adapter in that system to a
linecard installed in slot 0/opto of the second hub.
_____________________ Note _____________________
You cannot install a CCMLB linecard in slot opto only.
5.5.2.4 Connecting MC2 Cables in Standard Hub Mode Using Optical Converters
If you are using optical converters in an MC2 configuration, install an optical
converter module (CCMFB), with attached BN34R fiber optics cable, when
you install the CCMAB Memory Channel PCI adapter in each system in the
standard hub configuration. You should also connect the CCMAB Memory
Channel adapter to the optical converter with a BN39B-01 cable.
5–10 Setting Up the Memory Channel Cluster Interconnect
Now you need to:
•
•
•
Set the CCMLB linecard jumpers to support fiber optics
Connect the fiber optics cable to a CCMFB fiber optics converter module
Install the CCMFB fiber optics converter module for each fiber optics link
______________________
Note _______________________
Remember, if you have more than four fiber optics links, you need
two or more hubs. The CCMHB-BA hub has no linecards.
To set the CCMLB jumpers and install CCMFB optics converter modules
in an MC2 hub, follow these steps:
1. Remove the appropriate CCMLB linecard and set the linecard jumpers
to Fiber On (jumper pins 1 to 2) to support fiber optics. See Table 5–3.
2. Remove the CCMLB endplate and install the alternate endplate (with
the slot at the bottom).
3. Remove the hub bulkhead blanking plate from the appropriate hub slot.
Ensure that you observe the slot restrictions for the optical converter
modules. Also keep in mind that all linecards for one Memory Channel
interconnect must be in the same hub (see Section 5.4.)
4. Thread the BN34R fiber optics cable through the hub bulkhead slot.
The other end should be attached to a CCMFB optics converter in the
member system.
5. Thread the BN34R fiber optics cable through the slot near the bottom of
the endplate. Remove the cable tip protectors and insert the connectors
into the transceiver until they click into place. Secure the cable to the
module using the tie-wrap.
6. Install the CCMFB fiber optics converter in slot opto only, 0/opto,
1/opto, 2/opto, or 3/opto as appropriate.
7. Install a BN39B-01 1-meter link cable between the CCMFB optical
converter and the CCMLB linecard.
8. Repeat steps 1 through 7 for each CCMFB module to be installed.
5.6 Running Memory Channel Diagnostics
After the Memory Channel adapters, hubs, link cables, fiber optics
converters, and fiber optics cables have been installed, power up the systems
and run the Memory Channel diagnostics.
Setting Up the Memory Channel Cluster Interconnect 5–11
There are two console level Memory Channel diagnostics, mc_diag and
mc_cable:
•
The mc_diag diagnostic:
–
–
–
Tests the Memory Channel adapter(s) on the system running the
diagnostic.
Runs as part of the initialization sequence when the system is
powered up.
Runs on a standalone system or while connected to another system or
a hub with the link cable.
•
The mc_cable diagnostic:
–
Must be run on all systems in the cluster simultaneously (therefore,
all systems must be at the console prompt).
__________________ Caution
__________________
If you attempt to run mc_cable on one cluster member
while other members of the cluster are up, you may crash
the cluster.
–
–
Is designed to isolate problems to the Memory Channel adapter,
BC12N or BN39B link cables, hub linecards, fiber optics converters,
BN34R fiber optics cable, and, to some extent, to the hub.
Indicates data flow through the Memory Channel by response
messages.
–
–
–
Runs continuously until terminated with Ctrl/C.
Reports differences in connection state, not errors.
Can be run in standard or virtual hub mode.
When the console indicates a successful response from all other systems
being tested, the data flow through the Memory Channel hardware has
been completed and the test may be terminated by pressing Ctrl/C on each
system being tested.
Example 5–1 shows a sample output from node 1 of a standard hub
configuration. In this example, the test is started on node 1, then on node
0. The test must be terminated on each system.
5–12 Setting Up the Memory Channel Cluster Interconnect
Example 5–1: Running the mc_cable Test
>>> mc_cable
1
To exit MC_CABLE, type <Ctrl/C>
mca0 node id 1 is online
No response from node 0 on mca0
mcb0 node id 1 is online
No response from node 0 on mcb0
Response from node 0 on mca0
Response from node 0 on mcb0
mcb0 is offline
2
2
3
3
4
5
6
6
mca0 is offline
Ctrl/C
7
>>>
1
2
The mc_cable diagnostic is initiated on node 1.
Node 1 reports that mca0 is on line but has not communicated with the
Memory Channel adapter on node 0.
3
4
5
6
7
Node 1 reports that mcb0 is on line but has not communicated with the
Memory Channel adapter on node 0.
Memory Channel adapter mca0 has communicated with the adapter
on the other node.
Memory Channel adapter mcb0 has communicated with the adapter
on the other node.
Typing a Ctrl/C on node 0 terminates the test on that node and the
Memory Channel adapters on node 1 report off line.
A Ctrl/C on node 1 terminates the test.
Setting Up the Memory Channel Cluster Interconnect 5–13
6
Using Fibre Channel Storage
This chapter provides an overview of Fibre Channel, Fibre Channel
configuration examples, and information on Fibre Channel hardware
installation and configuration in a Tru64 UNIX or TruCluster Server Version
5.0A configuration.
The information includes how to determine the /dev/disk/dskn value that
corresponds to the Fibre Channel storagesets that have been set up as the
Tru64 UNIX boot disk, cluster root (/), cluster /usr, cluster /var, cluster
member boot, and quorum disks, and how to set up the bootdef_dev
console environment variable to facilitate Tru64 UNIX Version 5.0A and
TruCluster Server Version 5.0A installation.
______________________
Note _______________________
TruCluster Server Version 5.0A configurations require one or
more disks to hold the Tru64 UNIX operating system. The disk(s)
are either private disk(s) on the system that will become the first
cluster member, or disk(s) on a shared bus that the system can
access.
Whether or not you install the base operating system on a shared
disk, always shut down the cluster before booting the Tru64
UNIX disk.
This chapter discusses the following topics:
•
The procedure for Tru64 UNIX Version 5.0A or TruCluster Server
Version 5.0A installation using Fibre Channel disks (Section 6.1).
•
•
•
•
Fibre Channel overview (Section 6.2).
Example cluster configurations using Fibre Channel storage (Section 6.3).
A brief discussion of zoning and cascaded switches (Section 6.4).
The steps necessary to install the Fibre Channel hardware, base
operating system, and cluster software using disks accessible over the
Fibre Channel hardware (Section 6.5 through Section 6.10).
•
Changing the HSG80 from transparent to multiple-bus failover mode.
Using Fibre Channel Storage 6–1
•
A discussion on how you can use the emx manager (emxmgr) to display
the presence of Fibre Channel adapters, target ID mappings for a Fibre
Channel adapter, and the current Fibre Channel topology (Section 6.12).
6.1 Procedure for Installation Using Fibre Channel Disks
Use the following procedure to install Tru64 UNIX Version 5.0A or
TruCluster Server Version 5.0A using Fibre Channel disks. If you are
only installing Tru64 UNIX Version 5.0A, complete the first eight steps.
Complete all the steps for a TruCluster Server Version 5.0A installation.
Refer to the Tru64 UNIX Installation Guide, TruCluster Server Software
Installation manual, and other hardware manuals as appropriate for the
actual installation procedures.
1. Install the Fibre Channel switch (Section 6.5.1).
2. Install the KGPSA PCI-to-Fibre Channel host bus adapter
(Section 6.5.2).
3. Set up the HSG80 RAID array controllers for a fabric configuration
(Section 6.5.3).
4. Configure the HSG80 disks to be used for base operating system and
cluster installation. Be sure to set the identifier for each storage unit
you will use for operating system or cluster installation (Section 6.6.1).
5. Power on the system where you will install Tru64 UNIX Version 5.0A.
If this is a cluster installation, this system will also be the first cluster
member.
Use the console WWID manager (wwidmgr) utility to set the device
unit number for the Fibre Channel Tru64 UNIX Version 5.0A disk and
cluster member system boot disks (Section 6.6.2).
6. Use the WWID manager to set the bootdef_dev console environment
variable (Section 6.6.3).
7. Refer to the Tru64 UNIX Installation Guide and install the base
operating system from CD-ROM. The installation procedure will
recognize the disks for which you set the device unit number. Select the
disk you have chosen as the base operating system installation disk
from the list of disks provided (Section 6.7).
8. After the new kernel has booted to multi-user mode, shut down the
operting system and reset the bootdef_dev console environment
variable to provide multiple boot paths (Section 6.8).
Boot the system to multi-user mode and complete the operating system
installation.
9. Determine the /dev/disk/dskn values to be used for cluster
installation (Section 6.9).
6–2 Using Fibre Channel Storage
10. Use the disklabel utility to label the disks used to create the cluster
(Section 6.10).
11. Refer to the TruCluster Server Software Installation manual and install
the TruCluster Server software subsets and run the clu_create
command to create the first cluster member. Do not allow clu_create
to boot the system. Shut down the system to the console prompt
(Section 6.10).
12. Reset the bootdef_dev console environment variable to provide
multiple boot paths (Section 6.8). Boot the first cluster member.
13. Run clu_add_member on a cluster member system to add subsequent
cluster members.
Before you boot the system being added to the cluster, on the newly
added cluster member:
a. Use the wwidmgr utility with the -quickset option to set the
device unit number for the member system boot disk (Section 6.6.2).
b. Set the bootdef_dev console environment variable to one
reachable path to the member system boot disk (Section 6.6.3)
c. Boot genvmunix.
14. Create a new kernel, but do not reboot. Shut the system down and reset
the bootdef_dev console environment variable to provide multiple
boot paths to the member system boot disk (Section 6.8). Boot the new
cluster member system.
15. Repeat steps 13 and 14 to add other cluster member systems.
Consult the following documentation to assist you in Fibre Channel storage
configuration and administration:
•
•
•
•
•
•
KGPSA-BC PCI-to-Optical Fibre Channel Host Adapter User Guide
(AA-RF2J B-TE)
Compaq StorageWorks Fibre Channel Storage Switch User ’s Guide
(AA-RHBYA-TE)
Compaq StorageWorks SAN Switch 8 Installation and Hardware Guide
(EK-BCP24-IA/161355-001)
Compaq StorageWorks SAN Switch 16 Installation and Hardware Guide
(EK-BCP28-IA/161356-001)
Compaq StorageWorks HSG80 Array Controller ACS Version 8.5
Configuration Guide
Compaq StorageWorks HSG80 Array Controller ACS Version 8.5 CLI
Reference Guide
Using Fibre Channel Storage 6–3
•
Wwidmgr User ’s Manual
6.2 Fibre Channel Overview
Fibre Channel supports multiple protocols over the same physical interface.
Fibre Channel is primarily a protocol-independent transport medium;
therefore, it is independent of the function that it is used for.
The TruCluster Server uses the Fibre Channel Protocol (FCP) for SCSI to
use Fibre Channel as the physical interface.
Fibre Channel, with its serial transmission method overcomes the
limitations of parallel SCSI by providing:
•
•
•
•
Data rates of 100 MB/sec, 200 MB/sec, and 400 MB/sec
Support for multiple protocols
Better scalability
Improved reliability, serviceability, and availability
Fibre Channel uses an extremely high-transmit clock frequency to achieve
the high data rate. Using optical fibre transmission lines allows the
high-frequency information to be sent up to 40 km, the maximum distance
between transmitter and receiver. Copper transmission lines may be used
for shorter distances.
6.2.1 Basic Fibre Channel Terminology
The following list describes the basic Fibre Channel terminology:
F r a m e
All data is transferred in a packet of information
called a frame. A frame is limited to 2112 bytes. If
the information consists of more than 2112 bytes, it
is divided up into multiple frames.
The source and destination of a frame. A node
may be a computer system, a redundant array of
independent disks (RAID) array controller, or a disk
device. Each node has a 64-bit unique node name
(worldwide name) that is built into the node when it
is manufactured.
Nod e
N_P or t
Each node must have at least one Fibre Channel
port from which to send or receive data. This node
port is called an N_Port. Each port is assigned a
64-bit unique port name (worldwide name) when it
6–4 Using Fibre Channel Storage
is manufactured. An N_Port is connected directly
to another N_Port in a point-to-point topology. An
N_Port is connected to an F_Port in a fabric topology.
NL_P or t
In an arbitrated loop topology, information is routed
around a loop. The information is repeated by each
intermediate port until it reaches its destination.
The N_Port that contains this additional loop
functionality is an NL_Port.
A switch, or multiple interconnected switches,
that route frames between the originator node
(transmitter) and destination node (receiver).
F a br ic
F _P or t
The ports within the fabric (fabric port). This port is
called an F_port. Each F_port is assigned a 64-bit
unique node name and a 64-bit unique port name
when it is manufactured. Together, the node name
and port name make up the worldwide name.
F L_P or t
Lin k
An F_Port containing the loop functionality is called
an FL_Port.
The physical connection between an N_Port and
another N_Port or an N_Port and an F_Port. A
link consists of two connections, one to transmit
information and one to receive information. The
transmit connection on one node is the receive
connection on the node at the other end of the link.
A link may be optical fibre, coaxial cable, or shielded
twisted pair.
E_P or t
An expansion port on a switch used to make a
connection between two switches in the fabric.
6.2.2 Fibre Channel Topologies
Fibre Channel supports three different interconnect topologies:
•
•
Point-to-point (Section 6.2.2.1)
Fabric (Section 6.2.2.2)
Using Fibre Channel Storage 6–5
•
Arbitrated loop (Section 6.2.2.3)
______________________
Note _______________________
Although it is possible to interconnect an arbitrated loop with
fabric, hybrid configurations are not supported at the present
time, and therefore not discussed in this manual.
6.2.2.1 Point-to-Point
The point-to-point topology is the simplest Fibre Channel topology. In a
point-to-point topology, one N_Port is connected to another N_Port by a
single link.
Because all frames transmitted by one N_Port are received by the other
N_Port, and in the same order in which they were sent, frames require no
routing.
Figure 6–1 shows an example point-to-point topology.
Figure 6–1: Point-to-Point Topology
Node 2
N_Port
Node 1
Transmit
Receive
Transmit
Receive
N_Port
ZK-1534U-AI
6.2.2.2 Fabric
The fabric topology provides more connectivity than point-to-point topology.
The fabric topology can connect up to 224 ports.
The fabric examines the destination address in the frame header and routes
the frame to the destination node.
A fabric may consist of a single switch, or there may be several
interconnected switches (up to three interconnected switches is supported).
Each switch contains two or more fabric ports (F_Port) that are internally
6–6 Using Fibre Channel Storage
connected by the fabric switching function, which routes the frame from one
F_Port to another F_Port within the switch. Communication between two
switches is routed between two expansion ports (E_Ports).
When an N_Port is connected to an F_Port, the fabric is responsible for the
assignment of the Fibre Channel address to the N_Port attached to the
fabric. The fabric is also responsible for selecting the route a frame will take,
within the fabric, to be delivered to the destination.
When the fabric consists of multiple switches, the fabric can determine an
alternate route to ensure that a frame gets delivered to its destination.
Figure 6–2 shows an example fabric topology.
Figure 6–2: Fabric Topology
Node 1
Node 3
N_Port
Transmit
Receive
Transmit
Receive
Transmit
Receive
Transmit
Receive
F_Port
F_Port
N_Port
Fabric
Node 2
N_Port
Node 4
N_Port
Transmit
Receive
Transmit
Receive
Transmit Transmit
F_Port
F_Port
Receive
Receive
ZK-1536U-AI
6.2.2.3 Arbitrated Loop Topology
In an arbitrated loop topology, frames are routed around a loop created by
the links between the nodes.
In an arbitrated loop topology, a node port is called an NL_Port (node loop
port), and a fabric port is called an FL_Port (fabric loop port).
Figure 6–3 shows an example arbitrated loop topology.
Using Fibre Channel Storage 6–7
Figure 6–3: Arbitrated Loop Topology
Node 3
Node 1
Transmit
Receive
Transmit
NL_Port
NL_Port
Receive
Hub
Node 4
Node 2
Transmit
NL_Port
Receive
Transmit
NL_Port
Receive
ZK-1535U-AI
______________________
Note _______________________
The arbitrated loop topology is not supported by the Tru64 UNIX
or TruCluster Server products.
When support for Fibre Channel arbitrated loop is announced in
the TruCluster Server Software Product Description (SPD), the
technical update version of this information will be modified to
include arbitrated loop. The SPD will provide a pointer to the
technical update.
6.3 Example Fibre Channel Configurations Supported by
TruCluster Server
This section provides diagrams of some of the configurations supported by
TruCluster Server Version 5.0A. Diagrams are provided for both transparent
failover mode and multiple-bus failover mode.
6.3.1 Fibre Channel Cluster Configurations for Transparent Failover
Mode
With transparent failover mode:
•
The hosts do not know a failover has taken place (failover is transparent
to the hosts).
6–8 Using Fibre Channel Storage
•
•
The units are divided between an HSG80 port 1 and port 2.
If there are dual-redundant HSG80 controllers, controller A port 1 and
controller B port 2 are normally active; controller A port 1 and controller
B port 1 are normally passive.
•
If one controller fails, the other controller takes control and both its
ports are active.
Figure 6–4 shows a typical Fibre Channel cluster configuration using
transparent failover mode.
Figure 6–4: Fibre Channel Single Switch Transparent Failover
Configuration
Member
System
2
Member
System
1
KGPSA
KGPSA
DSGGA
RA8000/
ESA12000
ZK-1531U-AI
In transparent failover, units D00 through D99 are accessed through port 1
of both controllers. Units D100 through D199 are accessed through port 2 of
both HSG80 controllers (with the limit of a total of 128 storage units).
You cannot achieve a no-single-poit-of-failure (NSPOF) configuration using
transparent failover. The host cannot initiate failover, and if you lose a host
bus adapter, switch, or a cable, you lose the units behind at least one port.
Using Fibre Channel Storage 6–9
You can, however, add the hardware for a second bus (another KGPSA,
switch, and RA8000/ESA12000 with associated cabling) and use LSM to
mirror across the buses. However, because you cannot use LSM to mirror
the cluster root (/) file system, member boot partitions, the quorum disk,
or swap partitions you cannot obtain an NSPOF transparent failover
configuration, even though you have increased availability.
6.3.2 Fibre Channel Cluster Configurations for Multiple-Bus Failover
Mode
With multiple-bus failover:
•
The host controls the fail over by accessing units over a different path or
causing the access to the unit to be through the other HSG80 controller
(one controller does not fail over to the other controller of its own accord).
•
•
Each cluster member system has two or more KGPSA host bus adapters
(multiple paths to the storage units).
Normally, all available units (D0 through D199, with a limit of 128
storage units) are available at all host ports. Only one HSG80 controller
will be actively doing I/O for any particular storage unit.
However, both controllers can be forced active by preferring units to
one controller or the other (SET unit PREFERRED_PATH=THIS). By
balancing the preferred units, you can obtain the best I/O performance
using two controllers.
_____________________ Note _____________________
If you have preferred units, and the HSG80 controllers restart
because of an error condition or power failure, and one
controller restarts before the other controller, the HSG80
controller restarting first will take all the units, whether
they are preferred or not. When the other HSG80 controller
starts, it will not have access to the preferred units, and will
be inactive.
Therefore, you want to ensure that both HSG80 controllers
start at the same time under all circumstances to ensure
that the preferred units are seen by the controller they are
preferred to.
Figure 6–5, Figure 6–6, and Figure 6–7 show three different multiple-bus
NSPOF cluster configurations. The only difference is the fibre-optic cable
connection path between the switch and the HSG80 controller ports.
6–10 Using Fibre Channel Storage
If you consider the loss of a host bus adapter or switch, the configurations in
Figure 6–6 and Figure 6–7 will provide better throughput than Figure 6–5
because you still have access to both controllers. With Figure 6–5, if you lose
a host bus adapter or switch, you lose the use of a controller.
Figure 6–5: Multiple-Bus NSPOF Configuration Number 1
Member
System
1
Member
System
2
KGPSA
KGPSA
KGPSA
KGPSA
DSGGA
DSGGA
1 2 1 2
A B
RA8000/
ESA12000
ZK-1706U-AI
Using Fibre Channel Storage 6–11
Figure 6–6: Multiple-Bus NSPOF Configuration Number 2
Member
System
1
Member
System
2
KGPSA
KGPSA
KGPSA
KGPSA
DSGGA
DSGGA
1 2 1 2
A B
RA8000/
ESA12000
ZK-1707U-AI
6–12 Using Fibre Channel Storage
Figure 6–7: Multiple-Bus NSPOF Configuration Number 3
Member
System
1
Member
System
2
KGPSA
KGPSA
KGPSA
KGPSA
DSGGA
DSGGA
1 2 1 2
A B
RA8000/
ESA12000
ZK-1708U-AI
6.4 Zoning and Cascaded Switches
This section provides a brief overview of zoning and cascaded switches.
6.4.1 Zoning
A zone is a logical subset of the Fibre Channel devices connected to the
fabric. Zoning allows partitioning of resources for management and access
control. In some configurations, it may provide for more efficient use of
hardware resources by allowing one switch to serve multiple clusters or even
multiple operating systems.
Figure 6–8 provides an example configuration using zoning. This
configuration consists of two independent zones with each zone containing
an independent cluster.
Using Fibre Channel Storage 6–13
Figure 6–8: A Simple Zoned Configuration
Cluster 1
Member
System 1
KGPSA
Cluster 2
Member
System 1
Cluster 1
Member
System 2
Cluster 2
Member
System 2
KGPSA
KGPSA
KGPSA
1
2
3
4
5
6
7
8
DSGGA SWITCH
9 10 11 12 13 14 15 16
RA8000/
ESA12000
RA8000/
ESA12000
ZK-1709U-AI
______________________
Note _______________________
Only static zoning is supported; zones can only be changed when
all connected systems are shut down.
For information on setting up zoning, refer to the SAN Switch Zoning
documentation that is provided with the switch.
6.4.2 Cascaded Switches
Multiple switches may be connected to each other. When cascading switches,
a maximum of three switches is supported, with a maximum of two hops
between switches. The maximum hop length is 10 km longwave single-mode
or 500 meters shortwave multimode Fibre Channel cable.
6–14 Using Fibre Channel Storage
6.5 Installing and Configuring Fibre Channel Hardware
This section provides information about installing the Fibre Channel
hardware needed for a TruCluster Server configuration accessing storage
over the Fibre Channel.
Ensure that the member systems, the Fibre Channel switches, and the
HSG80 array controllers are placed within the lengths of the optical cables
you will be using.
______________________
Note _______________________
The maximum length of the optical cable between the KGPSA and
the switch or switch and the HSG80 array controller is 500 meters
via shortwave multimode Fibre Channel cable. The maximum
distance between switches in a cascaded switch configuration is
10 kilometers using longwave single-mode fibre.
6.5.1 Installing and Setting Up the Fibre Channel Switch
The Fibre Channel switches support up to 8 (DS-DSGGA-AA/DS-DSGGB-AA)
or 16 (DS-DSGGA-AB/DS-DSGGB-AB) full duplex 1.6025 Gbits/sec
ports. Each switch port can be connected to a KGPSA-BC or KGPSA-CA
PCI-to-Fibre Channel host bus adapter, an HSG80 array controller, or
another switch.
Each switch, except the DS-DSGGB-AB, has a front panel display and four
push buttons that you use to manage the switch. There are four menus
that allow you to configure, operate, obtain status, or test the switch. The
DS-DSGGB-AB is managed by way of a telnet session once the IP address
has been set (from a PC or terminal).
All switches have a 10Base-T Ethernet (RJ 45) port, and once the IP address
is set, the Ethernet connection allows you to manage the switch:
•
•
•
Remotely using a telnet TCP/IP connection
With Simple Network Management Protocol (SNMP)
Using Web management tools
______________________
Note _______________________
You have to set the IP address and subnet mask from the front
panel (or from a PC or terminal with the DS-DSGGB-AA) before
Using Fibre Channel Storage 6–15
you can manage the switch by way of a telnet session, SNMP, or
the Web.
The DSGGA switch has slots to accommodate up to four (DS-DSGGA-AA) or
eight (DS-DSGGA-AB) plug-in interface modules. Each interface module in
turn supports two Giga Bit Interface Converter modules (GBIC). The GBIC
module is the electrical-to-optical converter.
The shortwave GBIC supports 50-micron multimode fibre (MMF) using
the standard subscriber connector (SC) connector. The longwave GBIC
supports 9-micron, single-mode fibre optical cables. Only the 50-micron
MMF optical cable is supported between the host bus adapters and switches
or switches and HSG80 controllers for the TruCluster Server product.
Longwave single-mode fibre optical cables are supported between switches
in a cascaded switch configuration.
______________________
Note _______________________
If you need to install additional interface modules, do so before
placing the switch in a relatively inaccessible location because
you have to remove the top cover to install the interface modules.
The DS-DSGGB switch accommodates up to 8 (DS-DSGGB-AA) or 16
(DS-DSBBG-AB) GBIC modules.
6.5.1.1 Installing the Switch
Place the switch within 500 meters of the member systems (with KGPSA
PCI-to-Fibre Channel adapter) and the HSG80 array controllers.
You can mount the switches in a 48.7-cm (19-in) rackmount installation or
place the switch on a flat solid surface.
When you plan the switch location, ensure that you provide access to the
front of the switch. All cables plug into the front of the switch. Also, for
those switches with a control panel, the display and switches are on the
front of the switch.
For an installation, at a minimum, you have to:
1. Place the switch or install it in the rack.
2. Connect the Ethernet cable.
3. Connect the fibre-optic cables.
4. Connect power to the switch.
6–16 Using Fibre Channel Storage
5. Turn on the power. The switch runs a series of power-on self test
(POST) tests.
6. Set the switch IP address and subnet mask (see Section 6.5.1.2.2). You
can also set the switch name if desired (see Section 6.5.1.2.5). The
switch IP address and subnet mask must initially be set from the front
panel, except for the DS-DSGGB-AA 8-port Fibre Channel switch. In
this case you have to connect a PC or terminal to the switch. You must
use a telnet session to set the switch name.
7. Reboot the switch to enable the change in IP address and subnet mask
to take effect.
For more information on the individual switches, see the following
documentation:
•
•
•
Compaq StorageWorks Fibre Channel Storage Switch User ’s Guide
(AA-RHBYA-TE)
Compaq StorageWorks SAN Switch 8 Installation and Hardware Guide
(EK-BCP24-IA/161355-001)
Compaq StorageWorks SAN Switch 16 Installation and Hardware Guide
(EK-BCP28-IA/161356-001)
For more information on the DSGGB command set, see the Compaq
StorageWorks SAN Switch Fabric Operating System Management Guide
(EK-P20FF-GA/161358-001).
6.5.1.2 Managing the Fibre Channel Switches
You can manage the DS-DSGGA-AA, DS-DSGGA-AB, and DS-DSGGB-AB
switches, and obtain switch status from the front panel, by making a telnet
connection or by accessing the Web. The DS-DSGGB-AA does not have a
front panel, so you must use a telnet connection or use Web access.
Before you can make a telnet connection or access the switch via the Web,
you must assign an IP address and subnet mask to the Ethernet connection
using the front panel or from a PC or terminal (DS-DSGGB-AA).
6.5.1.2.1 Using the Switch Front Panel
The switch front panel consists of a display and four buttons. The display is
normally not active, but it lights up when any of the buttons are pressed.
The display has a timer. After approximately 30 seconds of inactivity, the
display will go out.
The four front panel buttons are:
Up — Upward triangle: Scrolls the menu up (which effectively moves down
the list of commands) or increases the value being displayed.
Using Fibre Channel Storage 6–17
Down — Downward triangle: Scrolls the menu down (which effectively
moves up the list of commands) or decreases the value being displayed.
______________________
Note _______________________
When the up or down buttons are used to increase or decrease
a numerical display, the number changes slowly at first, but
changes to fast mode if the button is held down. The maximum
number displayed is 255. An additional increment at a count of
255 resets the count to 0.
Tab/Esc — Leftward triangle: Allows you to tab through multiple optional
functions, for example, the fields in an IP address. You can use this button
to abort an entry, which takes you to the previous menu item. If pressed
repeatedly, the front panel display will turn off.
Enter — Rightward triangle: Causes the switch to accept the input you have
made and move to the next function.
6.5.1.2.2 Setting the Ethernet IP Address and Subnet Mask from the Front Panel
Before you telnet to the switch, you must connect the Ethernet cable and
then set the Ethernet IP address and subnet mask.
To use the front panel to set the Ethernet address and subnet mask, follow
these steps:
1. Press any of the switch front panel buttons to activate the display for
the top-level menu. If the Configuration Menu is not displayed, press
the down button repeatedly until it is displayed:
Select Menu:
Configuration Menu
____________________
Note _____________________
Pressing the down button selects the next lower top-level
menu. The top-level menus are:
Configuration Menu
Operation Menu
Status Menu
Test Menu
6–18 Using Fibre Channel Storage
2. Press Enter to display the first submenu item in the configuration
menu, Ethernet IP address:
Ethernet IP address:
10.00.00.10
--
The underline cursor denotes the selected address field.
Use the up or down button to increase or decrease the displayed number.
Use the Tab/Esc button to select the next field. Modify the address
fields until you have the address set correctly.
3. Use Enter to accept the value and step to the next submenu item
(Ethernet Submask), and then repeat step 2 to set the Ethernet subnet
mask.
4. Press Enter to accept the Ethernet subnet mask.
5. Press the Tab/Esc button repeatedly to get back to the top-level menu.
6. Press the down button to select the Operation Menu:
Select Menu:
Operation Menu
7. If the switch is operational, place the switch off line before rebooting
or you will lose any transmission in progress.
Press Enter to display the first submenu in the Operation Menu, Switch
Offline:
Operation Menu:
Switch Offline
8. Press the down button until the Reboot submenu item is displayed:
Operation Menu:
Reboot
9. Press Enter. You can change your mind and not reboot:
Reboot
Accept?
Yes No
10. Use the Tab/Esc button to select Yes. Press Enter to reboot the switch
and execute the POST tests.
____________________
Note _____________________
After changing any configuration menu settings, you must
reboot the switch for the change to take effect.
Refer to the switch documentation for information on other switch
configuration settings.
Using Fibre Channel Storage 6–19
6.5.1.2.3 Setting the DS-DSGGB-AA Ethernet IP Address and Subnet Mask from a
PC or Terminal
For the DS-DSGGB-AA switch, which does not have a front panel, you must
use a connection to a Windows 95/98/NT PC or video terminal to set the
Ethernet IP address and subnet mask.
To set the Ethernet IP address and subnet mask for the DS-DSGGB-AA
switch, follow these steps:
1. Connect the switch serial port to a terminal or PC COM port with a
standard serial cable with a DB9 connector. Note that the serial port is
only used for initial power-on self-test (POST) verification, IP address
configuration, or for resetting the factory/default settings.
2. If you are using a PC, start a remote communication program, for
example, HyperTerminal.
3. Set the port settings to 9600 bits per second, 8 bits per character, and no
parity.
4. Turn on power to the switch. The switch automatically connects to the
host and logs the user on to the switch as admin.
5. Enter the ipAddrSet command, then enter the IP address, subnet
mask, and gateway address (if necessary). For example:
admin> ipAddrSet
Ethernet IP Address [10.77.77.77]: 16.142.72.54
Ethernet Subnetmask [255.255.255.0]: Return
Fibre Channel IP Address [none]: Return
Fibre Channel Subnetmask [none]: Return
Gateway Address [none]: Return
admin> logout
6.5.1.2.4 Logging Into the Switch with a Telnet Connection
Before you telnet to a Fibre Channel switch, you must set the Ethernet IP
address and subnet mask.
______________________
Note _______________________
A serial port connection and a telnet session cannot both be active
(at the same time) with the DS-DSGGB-AA switch. The telnet
session takes precedence and the serial port session is aborted
when the telnet session is started.
You can use a telnet session to log in to the switch at one of three security
levels. The default user names, shown from lowest security level to highest
security level, are shown in Table 6–1.
6–20 Using Fibre Channel Storage
Table 6–1: Telnet Session Default User Names for Fibre Channel Switches
Description
DSGGA
DSGGB
other
n/a
Allows you to execute commands ending in
Show, such as dateShow and portShow.
user
user
Allows you to execute all commands ending in
Show, plus any commands from the help menu
that do not change the state of the switch, for
example, version and errDump. You can
change the passwords for all users up to and
including the current user ’s security level.
admin
n/a
admin
root
Provides access to all the commands that show
up in the help menu. Most switch administration
is done when logged in as admin.
Gives users access to an extensive command set
that can significantly alter system performance.
Root commands should only be used at the
request of Compaq customer service.
You can set the user names and passwords for users at or below the security
level of the present login level by executing the passwd command. Enter a
new user name (if desired) and a new password for the user.
______________________ Notes ______________________
Use Ctrl/H to correct typing errors.
Use the logout command to log out from any telnet connection.
6.5.1.2.5 Setting the Switch Name via Telnet Session
After you set the IP address and subnet mask, you can use a telnet session
to log in to the switch to complete other switch management functions or
monitor switch status. For example, if a system’s /etc/hosts file contains
an alias for the switch’s IP address, set the switch name to the alias. This
allows you to telnet to the switch name from that system. Telnet from a
system that has the IP address in its /etc/hosts file and set the switch
address as follows:
Using Fibre Channel Storage 6–21
# telnet 132.25.47.146
Return
User admin Return
Passwd
Return
:Admin> switchName fcsw1 Return
:Admin> switchName Return
fcsw1
:Admin>
______________________
Note _______________________
When you telnet to the switch the next time, the prompt will
include the switch name, for example:
fcsw1:Admin>
6.5.2 Installing and Configuring the KGPSA PCI-to-Fibre Channel
Adapter Module
The following sections discuss KGPSA installation and configuration.
6.5.2.1 Installing the KGPSA PCI-to-Fibre Channel Adapter Module
To install the KGPSA-BC or KGPSA-CA PCI-to-Fibre Channel adapter
modules follow these steps. For more information, see the following
documentation:
•
KGPSA-BC PCI-to-Optical Fibre Channel Host Adapter User Guide
(AA-RF2J B-TE)
•
64-Bit PCI-to-Fibre Channel Host Bus Adapter User Guide
(AA-RKPDA-TE/173941-001)
_____________________ Caution
_____________________
Static electricity can damage modules and electronic components.
We recommend using a grounded antistatic wrist strap and a
grounded work surface when handling modules.
1. If necessary, install the mounting bracket on the KGPSA-BC module.
Place the mounting bracket tabs on the component side of the board.
Insert the screws from the solder side of the board.
2. The KGPSA-BC should arrive with the gigabit link module (GLM)
installed. If not, close the GLM ejector mechanism. Then, align the
6–22 Using Fibre Channel Storage
GLM alignment pins, alignment tabs, and connector pins with the holes,
oval openings, and board socket. Press the GLM into place.
The KGPSA-CA does not use a GLM, it uses an embedded optical
shortwave multimode Fibre Channel interface.
3. Install the KGPSA in an open 32- or 64-bit PCI slot.
4. Insert the optical cable SC connectors into the KGPSA-BC GLM or
KGPSA-CA SC connectors. The SC connectors are keyed to prevent
their being plugged in incorrectly. Do not use unnecessary force. Do not
forget to remove the transparent plastic covering on the extremities
of the optical cable.
5. Connect the fibre-optic cables to the shortwave gigabit interface
converter modules (GBICs) in the DSGGA or DSGGB Fibre Channel
switch.
6.5.2.2 Setting the KGPSA-BC or KGPSA-CA to Run on a Fabric
The KGPSA host bus adapter defaults to the fabric mode, and can be used in
a fabric without taking any action. However, if you install a KGPSA that
has been used in the loop mode on another system, you will need to reformat
the KGPSA nonvolatile RAM (NVRAM) and configure it to run on a Fibre
Channel fabric configuration.
Use the wwidmgr utility to determine the mode of operation of the KGPSA
host bus adapter, and to set the mode if it needs changing (for example
from loop to fabric).
______________________ Notes ______________________
You must set the console to diagnostic mode to use the wwidmgr
utility for the following AlphaServer systems: AS1200, AS4x00,
AS8x00, GS60, GS60E, and GS140. Set the console to diagnostic
mode as follows:
P00>>> set mode diag
Console is in diagnostic mode
P00>>>
The console remains in wwid manager mode (or diagnostic mode
for the AS1200, AS4x00, AS8x00, GS60, GS60E, and GS140
systems), and you cannot boot until the system is re-initialized.
Use the init command or a system reset to re-initialize the
system after you have completed using the wwid manager.
If you try to boot the system and receive the following error,
initialize the console to get out of WWID manager mode, then
reboot:
Using Fibre Channel Storage 6–23
P00>>> boot
warning -- main memory zone is not free
P00>>> init
.
.
.
P00>>> boot
If you have initialized and booted the system, then shut down the
system and try to use the wwidmgr utility, you may be prevented
from doing so. If you receive the following error, initialize the
system and retry the wwidmgr command:
P00>>> wwidmgr -show adapter
wwidmgr available only prior to booting.
Reinit system and try again.
P00>>> init
.
.
.
P00>>> wwidmgr -show adapter
.
.
.
For more information on the wwidmgr utility, see the Wwidmgr
User ’s Manual which is on the Alpha Systems Firmware Update
CD-ROM in the DOC directory.
Use the worldwide ID manager to show all KGPSA adapters:
P00>>> wwidmgr -show adapter
Link is down.
item
adapter
WWN
Cur. Topo Next Topo
pga0.0.0.3.1 - Nvram read failed
[ 0] pga0.0.0.3.1
pgb0.0.0.4.0 - Nvram read failed
[ 1] pgb0.0.0.4.0
pgc0.0.0.5.1 - Nvram read failed.
1000-0000-c920-eda0
FABRIC
FABRIC
FABRIC
UNAVAIL
UNAVAIL
UNAVAIL
1000-0000-c920-da01
[ 2] pgc0.0.0.5.1
1000-0000-c920-cd9c
[9999] All of the above.
The Link is down message indicates that one of the adapters is not
available, probably due to its not being plugged into a switch. The warning
message Nvram read failed indicates that the KGPSA NVRAM has not
been initialized and formatted. The next topology will always be UNAVAIL
for the host bus adapter that has an unformatted NVRAM. Both messages
are benign and can be ignored for the fabric mode of operation. To correct
the Nvram read failed situation, use the wwidmgr -set adapter
command.
The previous display shows that all three KGPSA host bus adapters are set
for fabric topology as the current topology, the default. When operating in
a fabric, if the current topology is FABRIC, it does not matter if the next
topology is Unavail, or that the NVRAM is not formatted (Nvram read
failed).
6–24 Using Fibre Channel Storage
If, however, the current topology is LOOP, you have to change the topology to
FABRIC to operate in a fabric. You will never see the Nvram read failed
message if the current topology is LOOP. The NVRAM has to have been
formatted to change the current mode to LOOP.
Consider the case where the KGPSA current topology is LOOP as follows:
P00>>> wwidmgr -show adapter
item
adapter
WWN
Cur. Topo Next Topo
[ 0] pga0.0.0.3.1
[ 1] pgb0.0.0.4.0
1000-0000-c920-eda0
1000-0000-c920-da01
LOOP
LOOP
LOOP
LOOP
[9999] All of the above.
If the current topology for an adapter is LOOP, set an individual adapter to
FABRIC by using the item number for that adapter (for example, 0 or 1).
Use 9999 to set all adapters:
P00>>> wwidmgr -set adapter -item 9999 -topology fabric
Reformatting nvram
Reformatting nvram
Displaying the adapter information again will show the topology that the
adapters will assume after the next console initialization:
P00>>> wwidmgr -show adapter
item
[ 0]
[ 1]
adapter
pga0.0.0.4.1
pgb0.0.0.3.0
WWN
Cur. Topo Next Topo
1000-0000-c920-eda0
1000-0000-c920-da01
LOOP
LOOP
FABRIC
FABRIC
[9999] All of the above.
This display shows that the current topology for both KGPSA host bus
adapters is LOOP, but will be FABRIC after the next initialization.
A system initialization configures the KGPSAs to run on a fabric.
6.5.2.3 Obtaining the Worldwide Names of KGPSA Adapters
A worldwide name is a unique number assigned to a subsystem by the
Institute of Electrical and Electronics Engineers (IEEE) and set by the
manufacturer prior to shipping. The worldwide name assigned to a
subsystem never changes. You should obtain and record the worldwide
names of Fibre Channel components in case you need to verify their target
ID mappings in the operating system.
Fibre Channel devices have both a node name and a port name worldwide
name, both of which are 64-bit numbers. Most commands you use with Fibre
Channel only show the port name.
There are multiple ways to obtain the KGPSA port name worldwide name:
•
You can obtain the worldwide name from a label on the KGPSA module
before you install it.
•
You can use the show dev command as follows:
Using Fibre Channel Storage 6–25
P00>>> show dev
.
.
.
pga0.0.0.1.0
pgb0.0.0.2.0
PGA0
PGB0
WWN 1000-0000-c920-eda0
WWN 1000-0000-c920-da01
•
•
You can use the wwidmgr -show adapter command as follows:
P00>>> wwidmgr -show adapter
item
[ 0]
[ 1]
adapter
pga0.0.0.4.1
pgb0.0.0.3.0
WWN
Cur. Topo Next Topo
1000-0000-c920-eda0
1000-0000-c920-da01
FABRIC
FABRIC
FABRIC
FABRIC
[9999] All of the above.
If the operating system is installed, the worldwide name of a KGPSA
adapter is also displayed in the boot messages generated when the emx
driver attaches to the adapter when the adapter ’s host system boots. Or,
you can use the grep utility and obtain the worldwide name from the
/var/adm/messages file as follows:
# grep wwn /var/adm/messages
F/W Rev 2.20X2(1.12): wwn 1000-0000-c920-eda0
F/W Rev 2.20X2(1.12): wwn 1000-0000-c920-eda0
F/W Rev 2.20X2(1.12): wwn 1000-0000-c920-eda0
.
.
.
Record the worldwide name of each KGPSA adapter for later use.
6.5.3 Setting up the HSG80 Array Controller for Tru64 UNIX
Installation
This section covers setting up the HSG80 controller for operation with
Tru64 UNIX Version 5.0A and TruCluster Server Version 5.0A. For more
information on installing the HSG80, see the Compaq StorageWorks
HSG80 Array Controller ACS Version 8.5 Configuration Guide or Compaq
StorageWorks HSG80 Array Controller ACS Version 8.5 CLI Reference Guide.
To set up an HSG80 for TruCluster Server operation, follow these steps:
1. If not already installed, install the HSG80 controller(s) into the RA8000
or ESA12000 storage arrays.
2. If used, ensure that the external cache battery (ECB) is connected to the
controller cache module(s).
3. Install the fibre-optic cables between the KGPSA and the switch.
4. Set the power verification and addressing (PVA) ID. Use PVA ID 0 for
the enclosure that contains the HSG80 controller(s). Set the PVA ID to
2 and 3 on expansion enclosures (if present).
6–26 Using Fibre Channel Storage
____________________
Note _____________________
Do not use PVA ID 1:
With Port-Target-LUN (PTL) addressing, the PVA ID is used
to determine the target ID of the devices on ports 1 through
6 (the LUN is always zero). Valid target ID numbers are 0
through 15, excluding numbers 4 through 7. Target IDs 6
and 7 are reserved for the controller pair, and target IDs 4
and 5 are never used.
The enclosure with PVA ID 0 will contain devices with target
IDs 0 through 3; with PVA ID 2, target IDs 8 through 11;
with PVA ID 3, target IDs 12 through 15. Setting a PVA ID
of an enclosure to 1 would set target IDs to 4 through 7,
generating a conflict with the target IDs of the controllers.
5. Remove the program card ESD cover and insert the controller ’s program
card. Replace the ESD cover.
6. Install disks into storage shelves.
7. Connect a terminal to the maintenance port on one of the HSG80
controllers. You need a local connection to configure the controller for
the first time. The maintenance port supports serial communication
with the following default values:
•
•
•
•
9600 BPS
8 data bits
1 stop bit
No parity
8. Connect the RA8000 or ESA12000 to the power source and apply power.
____________________ Note _____________________
The KGPSA host bus adapters must be cabled to the switch,
with the system power applied before you turn power on to
the RA8000/ESA12000, in order for the HSG80 to see the
connection to the KGPSAs.
9. If an uninterruptible power supply (UPS) is used instead of the external
cache battery, to prevent the controller from periodically checking the
cache batteries after power is applied, enter the following command:
> set this CACHE_UPS
Using Fibre Channel Storage 6–27
____________________
Note _____________________
Setting the controller variable CACHE_UPS for one controller
sets it for both controllers.
10. From the maintenance terminal, use the show this and show other
commands to verify that controllers have the current firmware version.
See the Compaq StorageWorks HSG80 Array Controller ACS Version 8.5
CLI Reference Guide for information on upgrading the firmware.
11. To ensure proper operation of the HSG80 with Tru64 UNIX and
TruCluster Server, set the the controller values as follows:
set nofailover
clear cli
set multibus copy = this
clear cli
set this port_1_topology = offline
set this port_2_topology = offline
set other port_1_topology = offline
set other port_2_topology = offline
set this port_1_topology = fabric
set this port_2_topology = fabric
set other port_1_topology = fabric
set other port_2_topology = fabric
1
2
3
4
5
5
5
5
6
6
6
6
1
2
Remove any failover mode that may have been previously
configured.
Prevents the command line interpreter (CLI) from reporting a
misconfiguration error resulting from not having a failover mode
set.
3
Puts the controller pair into multiple-bus failover mode. Ensure
that you copy the configuration information from the controller
known to have a good array configuration.
__________________ Note ___________________
Use the command set failover copy =
this_controller to set transparent failover mode.
4
When the command is entered to set multiple-bus failover and copy
the configuration information to the other controller, the other
controller will restart. The restart may set off the audible alarm
(which is silenced by pressing the button on the EMU). The CLI
will display an event report, and continue reporting the condition
until cleared with the clear cli command.
6–28 Using Fibre Channel Storage
5
6
Takes the ports off line and resets the topology to prevent an error
message when setting the port topology.
Sets fabric as the switch topology.
12. Enter the show connection command as shown in Example 6–1
to determine the HSG80 connection names for the connections to
the KGPSA host bus adapters. For an RA8000/ESA12000 with
dual-redundant HSG80s in multiple-bus failover mode, there will be
four connections for each KGPSA in the cluster (as long as all four
HSG80 ports are connected to the same fabric).
For example, in a two-node cluster with two KGPSAs in each member
system, and an RA8000 or ESA12000 with dual-redundant HSG80s,
there will be 16 connections for the cluster. If you have other systems
or clusters connected to the switches in the fabric, there will be other
connections for the other systems. In Example 6–1, note that the !
(exclamation mark) is part of the connection name. The HOST_ID is the
KGPSA host name worldwide name and the ADAPTER_ID is the port
name worldwide name.
Example 6–1: Determine HSG80 Connection Names
HSG80 show connection
Connection
Name
Unit
Offset
Operating system
Controller Port
Address Status
!NEWCON49
TRU64_UNIX
THIS
2
230813
OL this
OL this
0
0
HOST_ID=1000-0000-C920-DA01 ADAPTER_ID=1000-0000-C920-DA01
!NEWCON50
!NEWCON51
!NEWCON52
!NEWCON53
!NEWCON54
!NEWCON55
!NEWCON56
TRU64_UNIX
THIS
1
230813
HOST_ID=1000-0000-C920-DA01
ADAPTER_ID=1000-0000-C920-DA01
TRU64_UNIX
THIS
2
230913
OL this
0
HOST_ID=1000-0000-C920-EDEB
ADAPTER_ID=1000-0000-C920-EDEB
TRU64_UNIX
THIS
1
230913
OL this
0
HOST_ID=1000-0000-C920-EDEB
ADAPTER_ID=1000-0000-C920-EDEB
TRU64_UNIX
OTHER
1
230913
OL other
0
HOST_ID=1000-0000-C920-EDEB
ADAPTER_ID=1000-0000-C920-EDEB
TRU64_UNIX
OTHER
1
230813
OL other
0
HOST_ID=1000-0000-C920-DA01
ADAPTER_ID=1000-0000-C920-DA01
TRU64_UNIX
OTHER
2
230913
OL other
0
HOST_ID=1000-0000-C920-EDEB
ADAPTER_ID=1000-0000-C920-EDEB
TRU64_UNIX
OTHER
2
230813
OL other
0
HOST_ID=1000-0000-C920-DA01
ADAPTER_ID=1000-0000-C920-DA01
.
.
.
!NEWCON61
!NEWCON62
TRU64_UNIX
HOST_ID=1000-0000-C921-086C
THIS
2
210513
OL this
0
0
ADAPTER_ID=1000-0000-C921-086C
TRU64_UNIX
OTHER
1
210513
OL other
Using Fibre Channel Storage 6–29
Example 6–1: Determine HSG80 Connection Names (cont.)
HOST_ID=1000-0000-C921-086C
ADAPTER_ID=1000-0000-C921-086C
!NEWCON63
!NEWCON64
!NEWCON65
TRU64_UNIX
HOST_ID=1000-0000-C921-0943
OTHER
1
offline
0
0
0
ADAPTER_ID=1000-0000-C921-0943
TRU64_UNIX
HOST_ID=1000-0000-C920-EDA0
OTHER
1
210413
OL other
ADAPTER_ID=1000-0000-C920-EDA0
TRU64_UNIX
OTHER
2
210513
OL other
HOST_ID=1000-0000-C921-086C
ADAPTER_ID=1000-0000-C921-086C
.
.
.
!NEWCON74
!NEWCON75
!NEWCON76
!NEWCON77
!NEWCON78
!NEWCON79
TRU64_UNIX
HOST_ID=1000-0000-C920-EDA0
THIS
2
210413
OL this
0
0
0
ADAPTER_ID=1000-0000-C920-EDA0
TRU64_UNIX
HOST_ID=1000-0000-C921-0A75
THIS
2
offline
ADAPTER_ID=1000-0000-C921-0A75
TRU64_UNIX
HOST_ID=1000-0000-C920-EDA0
THIS
1
210413
OL this
ADAPTER_ID=1000-0000-C920-EDA0
TRU64_UNIX
HOST_ID=1000-0000-C921-086C
THIS
1
210513
OL this
0
ADAPTER_ID=1000-0000-C921-086C
TRU64_UNIX
HOST_ID=1000-0000-C920-CB77
THIS
2
offline
0
ADAPTER_ID=1000-0000-C920-CB77
TRU64_UNIX
OTHER
1
offline
0
HOST_ID=1000-0000-C920-CB77
ADAPTER_ID=1000-0000-C920-CB77
.
.
.
____________________
Note _____________________
You can change the connection name with the HSG80 CLI
RENAME command. For example, assume that member
system pepicelli has two KGPSA Fibre Channel host
bus adapters, and that the worldwide name for KGPSA
pga is 1000-0000-C920-DA01. Example 6–1 shows that
the connections for pga are !NEWCON49, !NEWCON50,
!NEWCON54, and !NEWCON56. You could change the name of
!NEWCON49 to indicate that it is the first connection (of four)
to pga on member system pepicelli as follows:
HSG80> rename !NEWCON49 pep_pga_1
13. For each connection to your cluster, verify that the operating system is
TRU64_UNIX and the unit offset is 0. Search the show connection
display for the worldwide name of each of the KGPSA adapters in
6–30 Using Fibre Channel Storage
your cluster member systems. If the operating system and offsets are
incorrect, set them, then restart both controllers as follows:
HSG80> set !NEWCON49 unit_offset = 0
HSG80> set !NEWCON49 operating_system = TRU64_UNIX
HSG80> restart other
1
2
3
3
HSG80> restart this
.
.
.
HSG80> show connection
4
1
2
Set the relative offset for LUN numbering to 0. You can set the
unit_offset to nonzero values, but use caution. Make sure you
understand the impact.
Specify that the host environment connected to the Fibre
Channel port is TRU64_UNIX. You must change each connection
to TRU64_UNIX. This is very important. Failure to set this to
TRU64_UNIX will prevent your system from booting correctly,
recovering from run-time errors, or from booting at all. The default
operating system is Windows NT, and NT uses a different SCSI
dialect to talk to the HSG80 controller.
3
4
Restart both controllers to cause all changes to take effect.
Enter the show connection command once more and verify that
all connections have the offsets set to 0 and the operating system is
set to TRU64_UNIX.
____________________
Note _____________________
If the fibre-optic cables are not properly installed, there will
be inconsistencies in the connections shown.
14. Set up the storage sets as required for the applications to be used. An
example is provided in Section 6.6.1.
6.5.3.1 Obtaining the Worldwide Names of HSG80 Controller
The RA8000 or ESA12000 is assigned a worldwide name when the unit is
manufactured. The worldwide name (and checksum) of the unit appears
on a sticker placed above the controllers. The worldwide name ends in
zero (0), for example, 5000-1FE1-0000-0D60. You can also use the SHOW
THIS_CONTROLLER Array Controller Software (ACS) command.
For HSG80 controllers, the controller port IDs are derived from the
RA8000/ESA12000 worldwide name as follows:
Using Fibre Channel Storage 6–31
•
In a subsystem with two controllers in transparent failover mode, the
controller port IDs increment as follows:
–
–
Controller A and controller B, port 1 — worldwide name + 1
Controller A and controller B, port 2 — worldwide name + 2
For example, using the worldwide name of 5000-1FE1-0000-0D60, the
following port IDs are automatically assigned and shared between the
ports as a REPORTED PORT_ID on each port:
–
–
Controller A and controller B, port 1 — 5000-1FE1-0000-0D61
Controller A and controller B, port 2 — 5000-1FE1-0000-0D62
•
In a configuration with dual-redundant controllers in multiple-bus
failover mode, the controller port IDs increment as follows:
–
–
–
–
Controller A port 1 — worldwide name + 1
Controller A port 2 — worldwide name + 2
Controller B port 1 — worldwide name + 3
Controller B port 2 — worldwide name + 4
For example, using the worldwide name of 5000-1FE1-0000-0D60, the
following port IDs are automatically assigned and shared between the
ports as a REPORTED PORT_ID on each port:
–
–
–
–
Controller A port 1 — 5000-1FE1-0000-0D61
Controller A port 2 — 5000-1FE1-0000-0D62
Controller B port 1 — 5000-1FE1-0000-0D63
Controller B port 2 — 5000-1FE1-0000-0D64
Because the HSG80 controller ’s configuration information and worldwide
name is stored in nonvolatile random-access memory (NVRAM) on the
controller, there are different procedures for replacing HSG80 controllers
in an RA8000 or ESA12000:
•
If you replace one controller of a dual-redundant pair, the NVRAM
from the remaining controller retains the configuration information
(including worldwide name). When you install the replacement
controller, the existing controller transfers configuration information to
the replacement controller.
•
If you have to replace the HSG80 controller in a single controller
configuration, or if you must replace both HSG80 controllers in a
dual-redundant configuration simultaneously, you have two options:
6–32 Using Fibre Channel Storage
–
–
If the configuration has been saved to disk (with the
INITIALIZE DISKnnnn SAVE_CONFIGURATION or INITIALIZE
storageset-name SAVE_CONFIGURATION option), you can restore
it from disk with the CONFIGURATION RESTORE command.
If you have not saved the configuration to disk, but the label
containing the worldwide name and checksum is still intact, or you
have recorded the worldwide name and checksum (Section 6.5.3.1)
and other configuration information, you can use the command-line
interface (CLI) commands to configure the new controller and set the
worldwide name. Set the worldwide name as follows:
SET THIS NODEID=nnnn-nnnn-nnnn-nnnn checksum
6.6 Preparing to Install Tru64 UNIX and TruCluster Server
on Fibre Channel Storage
After the hardware has been installed and configured, there are preliminary
steps that must be completed before you install Tru64 UNIX and TruCluster
Server on Fibre Channel disks:
•
•
Configure HSG80 storagesets — In this document, example storagesets
are configured for both Tru64 UNIX and TruCluster Server on Fibre
Channel storage. Modify the storage configuration to meet your needs
(Section 6.6.1).
Set the device unit number — The device unit number is a subset of the
device name (as shown in a show device display). For example, in
the device name DKA100.1001.0.1.0, the device unit number is 100
(DKA100). The Fibre Channel worldwide name (often referred to as the
worldwide ID or WWID) is too long (64 bits) to be used as the device unit
number. Therefore, you set a device unit number that is an alias for the
Fibre Channel worldwide name (Section 6.6.2).
•
Set the bootdef_dev console environment variable — Before you
install the operating system (or cluster software), you must set the
bootdef_dev console environment variable to ensure that you boot
from the right disk (Section 6.6.3).
6.6.1 Configuring the HSG80 Storagesets
After the hardware has been installed and configured, storagesets must be
configured for software installation. The following disks/disk partitions are
needed for base operating system and cluster installation:
•
•
•
Tru64 UNIX disk
Cluster root (/)
Cluster /usr
Using Fibre Channel Storage 6–33
•
•
•
Cluster /var
Member boot disk (one for each cluster member system)
Quorum disk (if used)
If you are installing only the operating system, you need only the Tru64
UNIX disk (and of course any disks for applications). This document
assumes that both the base operating system and cluster software are to
be installed on Fibre Channel disks.
If you are installing a cluster, you need one or more disks to hold the Tru64
UNIX operating system. The disk(s) are either private disk(s) on the system
that will become the first cluster member, or disk(s) on a shared bus that
the system can access. Whether the Tru64 UNIX disk is on a private disk
or a shared disk, you should shut down the cluster before booting a cluster
member system standalone from the Tru64 UNIX disk.
An example configuration will show the procedure necessary to set up disks
for base operating system and cluster installation. Modify the procedure
according to your own disk needs. You can use any supported RAID level.
The example will be based on the use of four 4-GB disks used to create
two mirrorsets (RAID level 1) to provide reliability. The mirrorsets will be
partitioned to provide partitions of appropriate sizes. Disks 30200, 30300,
40000, and 40100 will be used for the mirrorsets.
Table 6–2 contains the necessary information to convert from the HSG80
unit numbers to /dev/disk/dskn and device names for the example
configuration. A blank table (Table A–1) is provided in Appendix A for use in
an actual installation.
One mirrorset, the BOOT-MIR mirrorset will be used for the Tru64 UNIX and
cluster member system boot disks. The other mirrorset, CROOT-MIR, will be
used for the cluster root (/), cluster /usr, cluster /var, and quorum disks.
To set up the example disks for operating system and cluster installation,
follow the steps in Example 6–2.
Example 6–2: Setting up the Mirrorset
HSG80> RUN CONFIG
1
Config Local Program Invoked
Config is building its table and determining what devices exist
on the system. Please be patient.
add disk DISK30200 3 2 0
add disk DISK30300 3 3 0
add disk DISK40000 4 0 0
add disk DISK40100 4 1 0
...
6–34 Using Fibre Channel Storage
Example 6–2: Setting up the Mirrorset (cont.)
Config - Normal Termination
HSG80> ADD MIRRORSET BOOT-MIR DISK 30200 40000
HSG80> ADD MIRRORSET CROOT-MIR DISK 30300 40100
HSG80> INITIALIZE BOOT-MIR
2
2
3
HSG80> INITIALIZE CROOT-MIR
HSG80> SHOW BOOT-MIR
3
4
Name
Storageset
Uses
Used by
--------------------------------------------------------------------
BOOT-MIR
mirrorset
DISK30200
DISK40000
Switches:
POLICY (for replacement) = BEST_PERFORMANCE
COPY (priority) = NORMAL
READ_SOURCE = LEAST_BUSY
MEMBERSHIP = 2, 2 members present
State:
UNKNOWN -- State only available when configured as a unit
Size: 8378028 blocks
HSG80> SHOW CROOT-MIR
4
Name
Storageset
Uses
Used by
--------------------------------------------------------------------
CROOT-MIR
mirrorset
DISK30300
DISK40100
Switches:
POLICY (for replacement) = BEST_PERFORMANCE
COPY (priority) = NORMAL
READ_SOURCE = LEAST_BUSY
MEMBERSHIP = 2, 2 members present
State:
UNKNOWN -- State only available when configured as a unit
Size: 8378028 blocks
HSG80> CREATE_PARTITION BOOT-MIR SIZE=25
HSG80> CREATE_PARTITION BOOT-MIR SIZE=25
HSG80> CREATE_PARTITION BOOT-MIR SIZE=LARGEST
HSG80> CREATE_PARTITION CROOT-MIR SIZE=5
HSG80> CREATE_PARTITION CROOT-MIR SIZE=15
HSG80> CREATE_PARTITION CROOT-MIR SIZE=40
HSG80> CREATE_PARTITION CROOT-MIR SIZE=LARGEST
HSG80> SHOW BOOT-MIR
5
5
5
6
6
6
6
7
Name
Storageset
Uses
Used by
---------------------------------------------------------------------
BOOT-MIR
mirrorset
DISK30200
DISK40000
Switches:
POLICY (for replacement) = BEST_PERFORMANCE
COPY (priority) = NORMAL
READ_SOURCE = LEAST_BUSY
MEMBERSHIP = 2, 2 members present
State:
UNKNOWN -- State only available when configured as a unit
Size:
8378028 blocks
Partitions:
Partition number
Size
Starting Block
Used by
---------------------------------------------------------------------
1
2
3
2094502
2094502
4189009
(
(
(
1072.38 MB)
1072.38 MB)
2144.77 MB)
0
2094507
4189014
8
9
10
Using Fibre Channel Storage 6–35
Example 6–2: Setting up the Mirrorset (cont.)
HSG80>
HSG80> SHOW CROOT-MIR
11
Name
Storageset
Uses
Used by
------------------------------------------------------------------------------
CROOT-MIR
mirrorset
DISK30300
DISK40100
Switches:
POLICY (for replacement) = BEST_PERFORMANCE
COPY (priority) = NORMAL
READ_SOURCE = LEAST_BUSY
MEMBERSHIP = 2, 2 members present
State:
UNKNOWN -- State only available when configured as a unit
Size:
Partitions:
Partition number
8378028 blocks
Size
Starting Block
Used by
---------------------------------------------------------------------
1
2
3
4
418896 (
214.47 MB)
643.42 MB)
1715.81 MB)
1715.81 MB)
0
418901
1675605
5026816
12
13
14
15
1256699
3351206
3351207
(
(
(
HSG80> ADD UNIT D131 BOOT-MIR PARTITION=1 DISABLE_ACCESS_PATH=ALL
HSG80> ADD UNIT D132 BOOT-MIR PARTITION=2 DISABLE_ACCESS_PATH=ALL
HSG80> ADD UNIT D133 BOOT-MIR PARTITION=3 DISABLE_ACCESS_PATH=ALL
HSG80> ADD UNIT D141 CROOT-MIR PARTITION=1 DISABLE_ACCESS_PATH=ALL
HSG80> ADD UNIT D142 CROOT-MIR PARTITION=2 DISABLE_ACCESS_PATH=ALL
HSG80> ADD UNIT D143 CROOT-MIR PARTITION=3 DISABLE_ACCESS_PATH=ALL
HSG80> ADD UNIT D144 CROOT-MIR PARTITION=4 DISABLE_ACCESS_PATH=ALL
16
HSG80> SET D131 IDENTIFIER=131
HSG80> SET D132 IDENTIFIER=132
HSG80> SET D133 IDENTIFIER=133
HSG80> SET D141 IDENTIFIER=141
HSG80> SET D142 IDENTIFIER=142
HSG80> SET D143 IDENTIFIER=143
HSG80> SET D144 IDENTIFIER=144
17
HSG80> set d131 ENABLE_ACCESS_PATH = !NEWCON49,!NEWCON50,!NEWCON51,!NEWCON52 18
HSG80> set d131 ENABLE_ACCESS_PATH = !NEWCON53,!NEWCON54,!NEWCON55,!NEWCON56
Warning 1000: Other host(s) in addition to the one(s) specified can still
access this unit. If you wish to enable ONLY the host(s)
specified, disable all access paths (DISABLE_ACCESS=ALL), then
again enable the ones specified
HSG80> set d131 ENABLE_ACCESS_PATH = !NEWCON61,!NEWCON62,!NEWCON64,!NEWCON65
Warning 1000: Other host(s) in addition to the one(s) specified can still
access this unit. If you wish to enable ONLY the host(s)
specified, disable all access paths (DISABLE_ACCESS=ALL), then
again enable the ones specified
HSG80> set d131 ENABLE_ACCESS_PATH = !NEWCON68,!NEWCON74,!NEWCON76,!NEWCON77
Warning 1000: Other host(s) in addition to the one(s) specified can still
access this unit. If you wish to enable ONLY the host(s)
specified, disable all access paths (DISABLE_ACCESS=ALL), then
again enable the ones specified
HSG80> set d132 ENABLE_ACCESS_PATH = !NEWCON49,!NEWCON50,!NEWCON51,!NEWCON52
.
.
.
HSG80> set d144 ENABLE_ACCESS_PATH = !NEWCON49,!NEWCON50,!NEWCON51,!NEWCON52
6–36 Using Fibre Channel Storage
Example 6–2: Setting up the Mirrorset (cont.)
HSG80> set d144 ENABLE_ACCESS_PATH = !NEWCON53,!NEWCON54,!NEWCON55,!NEWCON56
Warning 1000: Other host(s) in addition to the one(s) specified can still
access this unit. If you wish to enable ONLY the host(s)
specified, disable all access paths (DISABLE_ACCESS=ALL), then
again enable the ones specified
HSG80> set d144 ENABLE_ACCESS_PATH = !NEWCON61,!NEWCON62,!NEWCON64,!NEWCON65
Warning 1000: Other host(s) in addition to the one(s) specified can still
access this unit. If you wish to enable ONLY the host(s)
specified, disable all access paths (DISABLE_ACCESS=ALL), then
again enable the ones specified
HSG80> set d144 ENABLE_ACCESS_PATH = !NEWCON68,!NEWCON74,!NEWCON76,!NEWCON77
Warning 1000: Other host(s) in addition to the one(s) specified can still
access this unit. If you wish to enable ONLY the host(s)
specified, disable all access paths (DISABLE_ACCESS=ALL), then
again enable the ones specified
HSG80> show d131
19
LUN
Uses
Used by
------------------------------------------------------------------------------
D131
BOOT-MIR
(partition)
LUN ID:
6000-1FE1-0000-0D60-0009-8080-0434-002F
IDENTIFIER = 131
Switches:
RUN
NOWRITE_PROTECT
WRITEBACK_CACHE
READ_CACHE
READAHEAD_CACHE
MAXIMUM_CACHED_TRANSFER_SIZE = 32
Access:
!NEWCON49, !NEWCON50, !NEWCON51, !NEWCON52, !NEWCON53, !NEWCON54,
!NEWCON55, !NEWCON56, !NEWCON61, !NEWCON62, !NEWCON64, !NEWCON65,
!NEWCON68, !NEWCON74, !NEWCON76, !NEWCON77
State:
ONLINE to the other controller
NOPREFERRED_PATH
Size:
2094502 blocks
Geometry (C/H/S): ( 927 / 20 / 113 )
.
.
.
HSG80> show d144
19
LUN
Uses
Used by
------------------------------------------------------------------------------
D144
CROOT-MIR
(partition)
LUN ID:
6000-1FE1-0000-0D60-0009-8080-0434-0028
IDENTIFIER = 144
Switches:
RUN
NOWRITE_PROTECT
WRITEBACK_CACHE
READ_CACHE
READAHEAD_CACHE
MAXIMUM_CACHED_TRANSFER_SIZE = 32
Access:
!NEWCON49, !NEWCON50, !NEWCON51, !NEWCON52, !NEWCON53, !NEWCON54,
!NEWCON55, !NEWCON56, !NEWCON61, !NEWCON62, !NEWCON64, !NEWCON65,
!NEWCON68, !NEWCON74, !NEWCON76, !NEWCON77
State:
ONLINE to the other controller
NOPREFERRED_PATH
Size:
3351207 blocks
Geometry (C/H/S): ( 1483 / 20 / 113 )
Using Fibre Channel Storage 6–37
1
Use the CONFIG utility to configure the devices on the device side buses
and add them to the controller configuration. The CONFIG utility takes
about two minutes to complete. You can use the ADD DISK command to
add disk drives to the configuration manually.
2
3
Create the BOOT-MIR mirrorset using disks 30200 and 30300 and the
CROOT-MIR mirrorset using disks 40000 and 40100.
Initialize the BOOT-MIR and CROOT-MIR mirrorsets. If you want to set
any initialization switches, you must do so in this step. The BOOT-MIR
mirrorset will be used for the Tru64 UNIX and cluster member system
boot disks. The CROOT-MIR mirrorset will be used for the cluster root
(/), cluster /usr and cluster /var file systems, and the quorum disk.
4
5
Verify the mirrorset configuration and switches. Ensure that the
mirrorsets use the correct disks.
Create appropriate sized partitions in the BOOT-MIR mirrorset using
the percentage of the storageset that each partition will use. These
partitions will be used for the two member system boot disks (25 percent
or 1 GB each) and the Tru64 UNIX disk. For the last partition, the
controller assigns the largest free space available to the partition (which
will be close to 50 percent or 2 GB).
6
Create appropriate sized partitions in the CROOT-MIR mirrorset using
the percentage of the storageset that each partition will use. These
partitions will be used for the quorum disk (5 percent), cluster root
partition (15 percent), /usr (40 percent), and /var file systems. For
the last partition, /var, the controller assigns the largest free space
available to the partition (which will be close to 40 percent). See the
TruCluster Server Software Installation manual to obtain partition
sizes.
7
Verify the BOOT-MIR mirrorset partitions. Ensure that the partitions
are of the desired size. The partition number is in the first column,
followed by the partition size and starting block.
8
9
Partition for member system 1 boot disk.
Partition for member system 2 boot disk.
10 Partition for Tru64 UNIX operating system disk.
11 Verify the CROOT-MIR mirrorset partitions. Ensure that the partitions
are of the desired size. The partition number is in the first column,
followed by the partition size and starting block.
12 Partition for the quorum disk.
13 Partition for cluster root (/) filesystem.
6–38 Using Fibre Channel Storage
14 Partition for cluster /usr filesystem.
15 Partition for cluster /var filesystem.
16 Assign a unit number to each partition. When the unit is created by the
ADD UNIT command, disable access to all hosts. This allows selective
access in case there are other systems or clusters connected to the same
switch as our cluster.
Record the unit name of each partition with the intended use for that
partition (see Table 6–2).
17 Set the identifier for each storage unit. Use any number between 1 and
9999. The number you select for the storage unit shows up as the user
define identifier (UDID) in the wwidmgr -show wwid display. It will
be used by the WWID manager when setting the device unit number
and bootdef_dev console environment variable.
The identifier is also used with the hardware manager view devices
command (hwmgr -view devices) to locate the /dev/disk/dskn
value.
It will also show up during the Tru64 UNIX installation to allow you to
select the Tru64 UNIX installation disk.
____________________
Note _____________________
We recommend that you set the identifier for all Fibre
Channel storagesets. It provides a sure method of identifying
the storagesets. Make the identifiers unique numbers within
the domain (or within the cluster at a minimum). In other
words. do not use the same identifier on more than one
HSG80. The identifiers should be easily recognized. Ensure
that you record the identifiers (see Table 6–2).
18 Enable access to each unit for those hosts that you want to be able to
access this unit. Because access was initially disabled to all hosts, you
can ensure selective access to the units.
Use the connection name for each connection to the KGPSA host bus
adapter on the host for which you want access enabled. Many of the
connections used here are shown in Example 6–1.
19 Use the SHOW unit command (where unit is D131 through D133 and
D141 through 144 in the example) to verify the identifier and that
access to each unit is correct. Ensure that there is no connection to an
unwanted system. Record the identifier and worldwide name for later
use. Table 6–2 is a sample table filled in for the example. Table A–1 in
Appendix A is a blank table for your use in an actual installation. Note
Using Fibre Channel Storage 6–39
that at this point, even though the table is filled in, we do not yet know
the device names or dskn numbers.
Table 6–2: Converting Storageset Unit Numbers to Disk Names
Worldwide Name
User
Device Name
dskn
File System
or Disk
HSG80
Unit
Define
Identifier
(UDID)
dga131.1001.0.1.0 dsk17
dga132.1001.0.1.0 dsk16
dga133.1001.0.1.0 dsk15
131
132
133
141
142
143
144
Member 1 boot D131
disk
6000-1FE1-0000-0D60-
0009-8080-0434-002F
Member2 boot D132
disk
6000-1FE1-0000-0D60-
0009-8080-0434-0030
Tru64 UNIX
disk
D133
6000-1FE1-0000-0D60-
0009-8080-0434-002E
dsk21
Quorum disk
D141
6000-1FE1-0000-0D60-
0009-8080-0434-0029
N/Aa
dsk20
Cluster root (/) D142
6000-1FE1-0000-0D60-
0009-8080-0434-002A
N/Aa
/usr
dsk19
D143
6000-1FE1-0000-0D60-
0009-8080-0434-002B
N/Aa
/var
dsk18
D144
6000-1FE1-0000-0D60-
0009-8080-0434-0028
N/Aa
a
These units are not assigned an alias for the device unit number by the WWID manager command, therefore, they do not
get a device name and will not show up in a console show dev display.
6.6.2 Setting the Device Unit Number
Set the device unit number for the Fibre Channel disks to be used as the
Tru64 UNIX Version 5.0A installation disk and cluster member boot disks.
Setting the device unit number allows the installation scripts to recognize
a Fibre Channel disk. You have to set the device unit number because the
64-bit worldwide name is too large to be used as the device unit number.
When you set the device unit number, you set an alias for the device
worldwide name.
You use the WWID manager (wwidmgr) to define a device unit number that
is an alias for Fibre Channel devices. For instance, if DKA0 or DKA100 are
part of the device name seen in a show dev display, 0 or 100 is the device
unit number.
To set the device unit number for a Fibre Channel device, follow these steps:
1. Obtain the user define identifier (UDID) for the HSG80 storageset to
be used as the Tru64 UNIX Version 5.0A installation disk or cluster
member system boot disks. In the example in Table 6–2 the Tru64 UNIX
6–40 Using Fibre Channel Storage
disk is unit D133 with a UDID 133. The UDID for the cluster member 1
boot disk is 131, and the cluster member 2 boot disk is 132.
2. Use the wwidmgr -clear all command to clear the stored Fibre
Channel wwid1, wwid2, wwid3, wwid4, N1, N2, N3, and N4 console
environment variables. You want to start with all wwid<n> and N<n>
variables clear.
P00>>> wwidmgr -clear all
P00>>> show wwid*
wwid0
wwid1
wwid2
wwid3
P00>>> show n*
N1
N2
N3
N4
____________________
Note _____________________
The console only creates devices for which the wwid<n>
console environment variable has been set, and are accessible
through an HSG80 N_Port as specified by the N<n>
console environment variable also being set. These console
environment variables are set with the wwidmgr -quickset
or wwidmgr -set wwid commands. We use the wwidmgr
-quickset command later.
3. Use the wwidmgr -show wwid command to display the UDID and
worldwide names of all devices known to the console.
P00>>> wwidmgr -show wwid
[0] UDID:-1 WWID:01000010:6000-1fe1-0001-4770-0009-9171-3579-0008 (ev:none)
[1] UDID:-1 WWID:01000010:6000-1fe1-0001-4770-0009-9171-3579-0007 (ev:none)
[2] UDID:-1 WWID:01000010:6000-1fe1-0001-4770-0009-9171-3579-0009 (ev:none)
[3] UDID:-1 WWID:01000010:6000-1fe1-0001-4770-0009-9171-3579-000a (ev:none)
[4] UDID:-1 WWID:01000010:6000-1fe1-0001-4770-0009-9171-3579-000b (ev:none)
[5] UDID:-1 WWID:01000010:6000-1fe1-0001-4770-0009-9171-3579-000c (ev:none)
[6] UDID:-1 WWID:01000010:6000-1fe1-0001-4770-0009-9171-3579-000d (ev:none)
[7] UDID:-1 WWID:01000010:6000-1fe1-0001-4770-0009-9171-3579-000e (ev:none)
[8] UDID:-1 WWID:01000010:6000-1fe1-0001-4770-0009-9171-3579-000f (ev:none)
[9] UDID:-1 WWID:01000010:6000-1fe1-0001-4770-0009-9171-3579-0010 (ev:none)
[10] UDID:131 WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-002f (ev:none)
[11] UDID:132 WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-0030 (ev:none)
[12] UDID:133 WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-002e (ev:none)
[13] UDID:141 WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-0029 (ev:none)
[14] UDID:142 WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-002a (ev:none)
[15] UDID:143 WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-002b (ev:none)
Using Fibre Channel Storage 6–41
[16] UDID:144 WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-0028 (ev:none)
[17] UDID:-1 WWID:01000010:6000-1fe1-0000-0ca0-0009-8090-0708-002b (ev:none)
[18] UDID:-1 WWID:01000010:6000-1fe1-0000-0ca0-0009-8090-0708-002c (ev:none)
[19] UDID:-1 WWID:01000010:6000-1fe1-0000-0ca0-0009-8090-0708-002d (ev:none)
[20] UDID:-1 WWID:01000010:6000-1fe1-0000-0ca0-0009-8090-0708-002e (ev:none)
1
2
3
4
1
2
The number within the brackets ([ ]) is the item number of the
device shown on any particular line.
The UDID is assigned at the HSG80 with the set Dn IDENTIFIER
= xxx command, and is not used by the Tru64 UNIX operating
system, but may be set (as we have done with the SET D131
IDENTIFIER=131 group of commands). When the identifier is not
set at the HSG80, a value of -1 is displayed.
3
The worldwide name for the device. It is prefixed with the value
WWID:01000010:. The most significant 64 bits of the worldwide
name resembles the HSG80 worldwide name, and is assigned when
the unit is manufactured. The least significant 64 bits is a volume
serial number generated by the HSG80. You can use the HSG80
SHOW unit command to determine the worldwide name for each
storage unit (as shown in Example 6–2).
4
The console environment variable set for this worldwide name.
Only 4 wwid<n> console environment variables (wwid0, wwid1,
wwid2, and wwid3) can be set. The console show dev command
only shows those disk devices for which a wwid<n> console
environment variable has been set using the wwidmgr -quickset
or wwidmgr -set command. In this example, none of the wwid<n>
environment variables is set.
4. Look through the wwidmgr -show wwid display and locate the UDID
for the Tru64 UNIX disk (133) and each member system boot disks (131,
132) to ensure the storage unit is seen. As a second check, compare the
worldwide name values.
5. Example 6–3 shows the use of the wwidmgr command with the
-quickset option to define the UDID as the device unit number as an
alias for the worldwide name for each of the devices. The wwidmgr
-quickset utility sets the device unit number and also provides a
display of the device names and how the disk is reachable (reachability
display).
Example 6–3 shows:
•
The use of the wwidmgr -quickset command to set the device unit
number for the Tru64 UNIX Version 5.0A installation disk to 133,
the cluster member system boot disks to 131 (cluster member 1) and
6–42 Using Fibre Channel Storage
132 (cluster member 2). The device unit number is an alias for the
worldwide name for the storage unit.
•
The reachability part of the display provides the followng:
–
–
–
–
The worldwide name for the storage unit that is to be accessed
The new device name for the KGPSA
Whether access is available through a port
The HSG80 port (N_Port) that will be used to access the storage
unit
–
The connected column indicates the HSG80 controller ports
that will be used to access the storage units. The HSG80
controllers are in multiple-bus failover, so only one controller
is active.
Example 6–3: Using the wwidmgr quickset Command to Set Device Unit
Number
P00>>> wwidmgr -quickset -udid 133
Disk assignment and reachability after next initialization:
6000-1fe1-0000-0d60-0009-8080-0434-002e
via adapter:
pga0.0.0.1.0
pga0.0.0.1.0
pga0.0.0.1.0
pga0.0.0.1.0
pgb0.0.0.2.0
pgb0.0.0.2.0
pgb0.0.0.2.0
pgb0.0.0.2.0
via fc nport:
connected:
dga133.1001.0.1.0
dga133.1002.0.1.0
dga133.1003.0.1.0
dga133.1004.0.1.0
dgb133.1001.0.2.0
dgb133.1002.0.2.0
dgb133.1003.0.2.0
dgb133.1004.0.2.0
5000-1fe1-0000-0d64
5000-1fe1-0000-0d62
5000-1fe1-0000-0d63
5000-1fe1-0000-0d61
5000-1fe1-0000-0d64
5000-1fe1-0000-0d62
5000-1fe1-0000-0d63
5000-1fe1-0000-0d61
No
Yes
No
Yes
No
Yes
No
Yes
P00>>> wwidmgr -quickset -udid 131
Disk assignment and reachability after next initialization:
6000-1fe1-0000-0d60-0009-8080-0434-002e
via adapter:
pga0.0.0.1.0
pga0.0.0.1.0
pga0.0.0.1.0
pga0.0.0.1.0
pgb0.0.0.2.0
pgb0.0.0.2.0
pgb0.0.0.2.0
pgb0.0.0.2.0
via fc nport:
connected:
dga133.1001.0.1.0
dga133.1002.0.1.0
dga133.1003.0.1.0
dga133.1004.0.1.0
dgb133.1001.0.2.0
dgb133.1002.0.2.0
dgb133.1003.0.2.0
dgb133.1004.0.2.0
5000-1fe1-0000-0d64
5000-1fe1-0000-0d62
5000-1fe1-0000-0d63
5000-1fe1-0000-0d61
5000-1fe1-0000-0d64
5000-1fe1-0000-0d62
5000-1fe1-0000-0d63
5000-1fe1-0000-0d61
No
Yes
No
Yes
No
Yes
No
Yes
6000-1fe1-0000-0d60-0009-8080-0434-002f
via adapter:
via fc nport:
connected:
dga131.1001.0.1.0
dga131.1002.0.1.0
dga131.1003.0.1.0
dga131.1004.0.1.0
dgb131.1001.0.2.0
dgb131.1002.0.2.0
dgb131.1003.0.2.0
pga0.0.0.1.0
pga0.0.0.1.0
pga0.0.0.1.0
pga0.0.0.1.0
pgb0.0.0.2.0
pgb0.0.0.2.0
pgb0.0.0.2.0
5000-1fe1-0000-0d64
5000-1fe1-0000-0d62
5000-1fe1-0000-0d63
5000-1fe1-0000-0d61
5000-1fe1-0000-0d64
5000-1fe1-0000-0d62
5000-1fe1-0000-0d63
No
Yes
No
Yes
No
Yes
No
Using Fibre Channel Storage 6–43
Example 6–3: Using the wwidmgr quickset Command to Set Device Unit
Number (cont.)
dgb131.1004.0.2.0
pgb0.0.0.2.0
5000-1fe1-0000-0d61
Yes
P00>>> wwidmgr -quickset -udid 132
Disk assignment and reachability after next initialization:
6000-1fe1-0000-0d60-0009-8080-0434-002e
via adapter:
pga0.0.0.1.0
pga0.0.0.1.0
pga0.0.0.1.0
pga0.0.0.1.0
pgb0.0.0.2.0
pgb0.0.0.2.0
pgb0.0.0.2.0
pgb0.0.0.2.0
via fc nport:
connected:
dga133.1001.0.1.0
dga133.1002.0.1.0
dga133.1003.0.1.0
dga133.1004.0.1.0
dgb133.1001.0.2.0
dgb133.1002.0.2.0
dgb133.1003.0.2.0
dgb133.1004.0.2.0
5000-1fe1-0000-0d64
5000-1fe1-0000-0d62
5000-1fe1-0000-0d63
5000-1fe1-0000-0d61
5000-1fe1-0000-0d64
5000-1fe1-0000-0d62
5000-1fe1-0000-0d63
5000-1fe1-0000-0d61
No
Yes
No
Yes
No
Yes
No
Yes
6000-1fe1-0000-0d60-0009-8080-0434-002f
via adapter:
via fc nport:
connected:
dga131.1001.0.1.0
dga131.1002.0.1.0
dga131.1003.0.1.0
dga131.1004.0.1.0
dgb131.1001.0.2.0
dgb131.1002.0.2.0
dgb131.1003.0.2.0
dgb131.1004.0.2.0
pga0.0.0.1.0
pga0.0.0.1.0
pga0.0.0.1.0
pga0.0.0.1.0
pgb0.0.0.2.0
pgb0.0.0.2.0
pgb0.0.0.2.0
pgb0.0.0.2.0
5000-1fe1-0000-0d64
5000-1fe1-0000-0d62
5000-1fe1-0000-0d63
5000-1fe1-0000-0d61
5000-1fe1-0000-0d64
5000-1fe1-0000-0d62
5000-1fe1-0000-0d63
5000-1fe1-0000-0d61
No
Yes
No
Yes
No
Yes
No
Yes
6000-1fe1-0000-0d60-0009-8080-0434-0030
via adapter:
via fc nport:
connected:
dga132.1001.0.1.0
dga132.1002.0.1.0
dga132.1003.0.1.0
dga132.1004.0.1.0
dgb132.1001.0.2.0
dgb132.1002.0.2.0
dgb132.1003.0.2.0
dgb132.1004.0.2.0
P00>>> init
pga0.0.0.1.0
pga0.0.0.1.0
pga0.0.0.1.0
pga0.0.0.1.0
pgb0.0.0.2.0
pgb0.0.0.2.0
pgb0.0.0.2.0
pgb0.0.0.2.0
5000-1fe1-0000-0d64
5000-1fe1-0000-0d62
5000-1fe1-0000-0d63
5000-1fe1-0000-0d61
5000-1fe1-0000-0d64
5000-1fe1-0000-0d62
5000-1fe1-0000-0d63
5000-1fe1-0000-0d61
No
Yes
No
Yes
No
Yes
No
Yes
.
.
.
______________________ Notes ______________________
The wwidmgr -quickset command can take up to a minute to
complete on the AlphaServer 8x00, GS60, GS60E, and GS140
systems.
You must reinitialize the console after running the WWID
manager (wwidmgr), and keep in mind that the AS1200, AS4x00,
AS8x00, GS60, GS60E, and GS140 console is in diagnostic mode.
6–44 Using Fibre Channel Storage
The disks are not reachable and you cannot boot until after the
system is initialized.
Note, that in the reachability portion of the display, the storagesets are
reachable from KGPSA dga through two HSG80 ports and from KGPSA dgb
through two HSG80 ports. Also note, that the device unit numbers, the
alias for the worldwide name of the disk device, has been set for the KGPSA
for each HSG80 port. The device names have also been set for the cluster
member boot disks. Record the device names.
The wwidmgr -quickset command provides a reachability display
(equivalent to execution of the wwidmgr -reachability command). The
devices shown in the reachability display are available for booting and the
setting of the bootdef_dev console environment variable during normal
console mode.
If you execute the show wwid* console command now, it would show that
the environment variable wwidn is set for the three boot disks. Also, the
show n* command shows that the units are accessible through 4 HSG80
N_Ports as follows:
P00>>> show wwid*
wwid0 133 1 WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-002e
wwid1 131 1 WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-002f
wwid2 132 1 WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-0030
wwid3
P00>>> show n*
N1
N2
N3
N4
50001fe100000d64
50001fe100000d62
50001fe100000d63
50001fe100000d61
Example 6–4 provides sample device names as displayed by the show dev
command after using the wwidmgr -quickset command to set the device
unit numbers.
Example 6–4: Sample Fibre Channel Device Names
P00>>> show dev
dga131.1001.0.1.0
dga131.1002.0.1.0
dga131.1003.0.1.0
dga131.1004.0.1.0
dga132.1001.0.1.0
dga132.1002.0.1.0
dga132.1003.0.1.0
dga132.1004.0.1.0
dga133.1001.0.1.0
dga133.1002.0.1.0
dga133.1003.0.1.0
dga133.1004.0.1.0
dgb131.1001.0.2.0
dgb131.1002.0.2.0
dgb131.1003.0.2.0
dgb131.1004.0.2.0
$1$DGA131
$1$DGA131
$1$DGA131
$1$DGA131
$1$DGA132
$1$DGA132
$1$DGA132
$1$DGA132
$1$DGA133
$1$DGA133
$1$DGA133
$1$DGA133
$1$DGA131
$1$DGA131
$1$DGA131
$1$DGA131
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
Using Fibre Channel Storage 6–45
Example 6–4: Sample Fibre Channel Device Names (cont.)
dgb132.1001.0.2.0
dgb132.1002.0.2.0
dgb132.1003.0.2.0
dgb132.1004.0.2.0
dgb133.1001.0.2.0
dgb133.1002.0.2.0
dgb133.1003.0.2.0
dgb133.1004.0.2.0
dka0.0.0.1.1
$1$DGA132
$1$DGA132
$1$DGA132
$1$DGA132
$1$DGA133
$1$DGA133
$1$DGA133
$1$DGA133
DKA0
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
HSG80 V8.5F
COMPAQ BB00911CA0 3B05
COMPAQ CDR-8435 0013
dqa0.0.0.15.0
DQA0
dva0.0.0.1000.0
ewa0.0.0.5.1
DVA0
EWA0
08-00-2B-C4-61-11
pga0.0.0.1.0
pgb0.0.0.2.0
pka0.7.0.1.1
PGA0
PGB0
PKA0
WWN 1000-0000-c920-eda0
WWN 1000-0000-c920-da01
SCSI Bus ID 7 5.57
______________________
Note _______________________
The only Fibre Channel devices displayed by the console show
dev command are those devices that have been assigned to a
wwid<n> environment variable.
Before you start the Tru64 UNIX installation, you must set the
bootdef_dev console environment variable.
6.6.3 Setting the bootdef_dev Console Environment Variable
When booting from Fibre Channel devices, you must set the bootdef_dev
console environment variable to ensure that the installation procedure is
able to boot the system after building the new kernel.
______________________ Notes ______________________
The bootdef_dev environment variable values must point to
the same HSG80.
After the base operating system has been installed, or after
cluster software has been installed for a cluster member system,
set the bootdef_dev console environment variable again to
provide multiple boot paths (see Section 6.8).
It would do no good to set bootdef_dev for multiple boot paths
now because the installation procedure overwrites the variable.
The following procedure is used for:
•
The initial Tru64 UNIX installation before booting from CD-ROM.
6–46 Using Fibre Channel Storage
•
After a cluster member has been added to the cluster with
clu_add_member (but before the member system is booted).
_____________________ Note _____________________
You do not use this procedure after using clu_create to
create the first cluster member. Before booting the first cluster
member, you reset the bootdef_dev console environment
variable to multiple boot paths.
To set the bootdef_dev console environment variable when booting from a
Fibre Channel device, follow these steps:
1. Obtain the device name for the Fibre Channel units that you will boot
from. Ensure that you choose the correct device name for the entity you
are booting (Tru64 UNIX, cluster member system 2, and so on). They
show up in the reachability display as shown in Example 6–3 with a Yes
under the connected column. You can also use the wwidmgr -show
reachability command to determine reachability. Example 6–4
provides the display for a show dev command, which shows the device
names of devices that may be assigned to the bootdef_dev console
environment variable. Example 6–3 and Example 6–4 show that the
following device names can be used in the bootdef_dev console
environment variable as possible boot devices:
•
•
•
•
•
•
•
•
•
•
•
•
dga131.1002.0.1.0
dga131.1004.0.1.0
dga132.1002.0.1.0
dga132.1004.0.1.0
dga133.1002.0.1.0
dga133.1004.0.1.0
dgb131.1002.0.2.0
dgb131.1004.0.2.0
dgb132.1002.0.2.0
dgb132.1004.0.2.0
dgb133.1002.0.2.0
dgb133.1004.0.2.0
Note that each of the storage units are reachable through four different
paths, two for each host bus adapter (the Yes in the connected column).
Using Fibre Channel Storage 6–47
2. Set the bootdef_dev console environment variable to one of the
boot path(s) that show up as connected. Ensure that you set the
bootdef_dev variable appropriately for the system and boot disk. For
the example disk configuration, set bootdef_dev as follows:
•
On the system where you are installing the Tru64 UNIX operating
system (which will also be the first cluster member):
P00>>> set bootdef_dev dga133.1002.0.1.0
•
For a system that has just been set up as the second (or subsequent)
cluster member system:
P00>>> set bootdef_dev dga132.1002.0.1.0
3. You must initialize the system to use any of the device names in the
bootdef_dev variable:
P00>>> init
.
.
.
After the initialization, the bootdef_dev will show up as follows:
P00>>> show bootdef_dev
bootdef_dev
dga133.1002.0.1.0
or:
P00>>> show bootdef_dev
bootdef_dev
dga132.1002.0.1.0
Now you are ready to install the Tru64 UNIX operating system, or boot
genvmunix on a cluster member system that has been added to the cluster
with clu_add_member.
6.7 Install the Base Operating System
After reading the TruCluster Server Software Installation manual, and
using the Tru64 UNIX Installation Guide as a reference, boot from the
CD-ROM and perform a full installation of the Tru64 UNIX Version 5.0A
operating system.
When the installation procedure displays the list of disks available for
operating system installation as shown here, look for the identifier in
the Location column. Verify the identifier from the table you have been
preparing (see Table 6–2).
To visually locate a disk, enter "ping <disk>",
where <disk> is the device name (for example, dsk0) of the disk you
want to locate. If that disk has a visible indicator light, it will
blink until you are ready to continue.
6–48 Using Fibre Channel Storage
Device
Name
dsk0
dsk15
dsk16
dsk17
Size Controller Disk
in GB Type
4.0 SCSI
1.0 SCSI
1.0 SCSI
2.0 SCSI
Model
RZ2CA-LA
HSG80
HSG80
HSG80
Location
1)
2)
3)
4)
bus-0-targ-0-lun-0
IDENTIFIER=133
IDENTIFIER=132
IDENTIFIER=131
If you flash the light on a storage unit (logical disk) that is a mirrorset,
stripeset, or RAIDset, the lights on all disks in the storageset will blink.
Record the /dev/disk/dskn value (dsk15) for the Tru64 UNIX disk that
matches the UDID (133) (Table 6–2).
Complete the installation, following the instructions in the Tru64 UNIX
Installation Guide.
After the installation is complete, reset the bootdef_dev console
environment variable to provide multiple boot paths (see Section 6.8).
6.8 Resetting the bootdef_dev Console Environment
Variable
If you set the bootdef_dev console environment variable to multiple paths
in Section 6.6.3, the base operating system installation, clu_create, or
clu_add_member procedures modify the variable and you should reset it to
provide multiple boot paths.
To reset the bootdef_dev console environment variable, follow these steps:
1. Obtain the device name and worldwide name for the Fibre Channel
units that you will boot from (see Table 6–2). Ensure that you choose
the correct device name for the entity you are booting (Tru64 UNIX,
cluster member system 2, and so on).
2. Check the reachability display (Example 6–3) provided by the wwidmgr
-quickset or the wwidmgr -reachability commands for the device
names that can access the storage unit you are booting from. Check the
show dev command output to ensure the device name may be assigned
to the bootdef_dev console environment variable.
____________________ Notes
____________________
You should choose device names that show up as both Yes
and No in the reachability display connected column. Keep
in mind, that for multiple-bus failover, only one controller is
normally active for a storage unit. You must ensure that the
unit is reachable if the controllers have failed over.
If you have multiple Fibre Channel host bus adapters, you
should use device names for at least two host bus adapters.
Using Fibre Channel Storage 6–49
For example, to ensure that you have a connected boot path
in case of a failed host bus adapter or controller failover,
choose device names for multiple host bus adapters and each
controller port. For example, if you use the reachability
display shown in Example 6–3, you could choose the
following device names when setting the bootdef_dev
console environment variable:
dga133.1001.0.1.0
dga133.1004.0.1.0
dgb133.1002.0.2.0
dgb133.1003.0.2.0
1
2
3
4
1
2
3
4
Path from host bus adapter A to controller B port 2
Path from host bus adapter A to controller A port 1
Path from host bus adapter B to controller A port 2
Path from host bus adapter B to controller B port 1
You can set units preferred to a specific controller, in which
case both controllers will be active.
3. Set the bootdef_dev console environment variable to a comma
separated list of several of the boot path(s) that show up as connected
in the reachability display (wwidmgr -quickset or wwidmgr -show
reachability). You must initialize the system to use any of the device
names in the bootdef_dev variable as follows:
•
For the base operating system:
P00>>> set bootdef_dev \
dga133.1001.0.1.0,dga133.1002.0.2.0,\
dgb133.1001.0.1.0,dgb133.1002.0.2.0
POO>>> init
.
.
.
•
For member system 1 boot disk:
P00>>> set bootdef_dev \
dga131.1001.0.1.0,dga131.1002.0.2.0,\
dgb131.1001.0.1.0,dgb131.1002.0.2.0
POO>>> init
.
.
.
6–50 Using Fibre Channel Storage
•
For member system 2 boot disk:
P00>>> set bootdef_dev \
dga132.1001.0.1.0,dga132.1002.0.2.0,\
dgb132.1001.0.1.0,dgb132.1002.0.2.0
POO>>> init
.
.
.
______________________
Note _______________________
The console system reference manual (SRM) software guarantees
that you can set the bootdef_dev console environment variable
to a minimum of four device names. You may be able to set it to
five, but four is all that is guaranteed.
6.9 Determining /dev/disk/dskn to Use for a Cluster
Installation
Before you can install the TruCluster Server software, you must determine
which /dev/disk/dskn to use for the various TruCluster Server disks.
To determine the /dev/disk/dskn to use for the cluster disks, follow these
steps:
1. With the Tru64 UNIX Version 5.0A operating system at single-user or
multi-user mode, use the hardware manager (hwmgr) utility with the
-view devices option to display all devices on the system. Use the
grep utility to search for any items with the IDENTIFIER qualifier.
# hwmgr -view dev | grep IDENTIFIER
HWID: Device Name
Mfg
Model
Location
-----------------------------------------------------------------------
62: /dev/disk/dsk15c
63: /dev/disk/dsk16c
64: /dev/disk/dsk17c
65: /dev/disk/dsk18c
66: /dev/disk/dsk19c
67: /dev/disk/dsk20c
68: /dev/disk/dsk21c
DEC
DEC
DEC
DEC
DEC
DEC
DEC
HSG80
HSG80
HSG80
HSG80
HSG80
HSG80
HSG80
IDENTIFIER=133
IDENTIFIER=132
IDENTIFIER=131
IDENTIFIER=141
IDENTIFIER=142
IDENTIFIER=143
IDENTIFIER=144
____________________
Note _____________________
If you know that you have set the UDID for a large number
of disks, you can also grep for the UDID:
# hwmgr -view dev | grep IDENTIFIER | grep 131
If you have not set the UDID, you can use hwmgr to determine
the /dev/disk/dskn name by using the hardware manager
to display device attributes and searching for the storage
unit worldwide name as follows:
Using Fibre Channel Storage 6–51
# hwmgr -get attribute -a name -a dev_base_name
| more
Use the more search utility (/) to search for the worldwide
name of the storageset you have set up for the particular
disk in question. The following example shows the format of
the command output:
# hwmgr -get attribute -a name -a dev_base_name
1:
name = Compaq AlphaServer ES40
2:
name = CPU0
.
.
.
62:
name = SCSI-WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-002e
dev_base_name = dsk15
63:
name = SCSI-WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-0030
dev_base_name = dsk16
64:
name = SCSI-WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-002f
dev_base_name = dsk17
65:
name = SCSI-WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-0028
dev_base_name = dsk18
66:
name = SCSI-WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-002b
dev_base_name = dsk19
67:
name = SCSI-WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-002a
dev_base_name = dsk20
68:
name = SCSI-WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-0029
dev_base_name = dsk21
69:
name = SCSI-WWID:0710002c:"COMPAQ CDR-8435 :d05b003t00000l00000"
dev_base_name = cdrom0
.
.
.
For more information on the hardware manager (hwmgr),
see hwmgr(8).
2. Search the display for the UDIDs (or worldwide names) for each of the
cluster installation disks and record the /dev/disk/dskn values.
If you used the grep utility to search for a specific UDID, for example
hwmgr -view dev | grep "IDENTIFIER=131" repeat the command
to determine the /dev/disk/dskn for each of the remaining cluster
disks. Record the information for use when you install the cluster
software.
6–52 Using Fibre Channel Storage
6.10 Installing the TruCluster Server Software
This section covers the Fibre Channel specific procedures you need to
execute before running clu_create to create the first cluster member
or clu_add_member to add subsequent cluster members. It also
covers the procedure you need to execute after running clu_create or
clu_add_member before you boot the new cluster member into the cluster.
Use the TruCluster Server Software Installation procedures in conjunction
with this manual for the TruCluster Server software installation.
To install the TruCluster Server software, follow these steps:
1. On the system you installed the Tru64 UNIX operating system on,
boot the system, and before you install the TruCluster Server software
subsets, determine the /dev/disk/dskn values to use for cluster
installation (see Section 6.9).
2. Initialize disklabels for all disks needed to create the cluster. For the
example disks we are using, it is disks dsk18 (/var), dsk19 (/usr),
dsk20 [cluster root (/)], and dsk21 (Quorum). For instance:
# disklabel -rw dsk20 HSG80
3. Install the TruCluster Server software subsets and run the clu_create
command to create the first cluster member using the procedures in the
TruCluster Server Software Installation manual. When clu_create
terminates, do not reboot the system. Shut down the system and reset
the bootdef_dev console environment variable to provide multiple boot
paths to the member system boot disk before booting (see Section 6.8).
Boot the first cluster member.
4. On the system you installed the Tru64 UNIX operating system on, run
clu_add_member to add subsequent cluster members.
____________________
Note _____________________
The system you installed the Tru64 UNIX operating system
on is already enabled to access all the member system boot
disks. If you use another cluster member system, you need to
use the wwidmgr -quickset command to set up the paths
to the member system boot disk.
Before you boot the system being added to the cluster, on the newly
added cluster member:
a. Use the wwidmgr utility with the -quickset option to set the
device unit number for the member system boot disk. For member
system 2 in the example configuration, it is the storage unit with
UDID 132 (See Table 6–2):
Using Fibre Channel Storage 6–53
P00>>> wwidmgr -quickset -udid 132
Disk assignment and reachability after next initialization:
6000-1fe1-0000-0d60-0009-8080-0434-0030
via adapter:
pga0.0.0.1.0
pga0.0.0.1.0
pga0.0.0.1.0
pga0.0.0.1.0
pgb0.0.0.2.0
pgb0.0.0.2.0
pgb0.0.0.2.0
pgb0.0.0.2.0
via fc nport:
connected:
dga132.1001.0.1.0
dga132.1002.0.1.0
dga132.1003.0.1.0
dga132.1004.0.1.0
dgb132.1001.0.2.0
dgb132.1002.0.2.0
dgb132.1003.0.2.0
dgb132.1004.0.2.0
5000-1fe1-0000-0d64
5000-1fe1-0000-0d62
5000-1fe1-0000-0d63
5000-1fe1-0000-0d61
5000-1fe1-0000-0d64
5000-1fe1-0000-0d62
5000-1fe1-0000-0d63
5000-1fe1-0000-0d61
No
Yes
No
Yes
No
Yes
No
Yes
b. Set the bootdef_dev console environment variable to one
reachable path (Yes in the connected column) to the member
system boot disk (See Section 6.6.3):
P00>>> set bootdef_dev dga132.1002.0.1.0
c. Boot genvmunix on the newly added cluster member system.
5. After the new kernel is built, do not reboot the new cluster member
system. Shut down the system and reset the bootdef_dev console
environment variable to provide multiple boot paths to the member
system boot disk before booting (see Section 6.8).
6. Repeat steps 4 and 5 for other cluster member systems.
6.11 Changing the HSG80 from Transparent to Multiple-Bus
Failover Mode
You may be are using transparent failover mode with Tru64 UNIX Version
5.0A and TruCluster Server Version 5.0A and want to take advantage of the
ability to create a no-single-point-of-failure (NSPOF) configuration, and the
availability that multiple-bus failover provides over transparent failover
mode.
If you are upgrading from Tru64 UNIX Version 4.0F or Version 4.0G and
TruCluster Software Products Version 1.6 to Tru64 UNIX Version 5.0A and
TruCluster Server Version 5.0A you may want to change from transparent
failover to multiple-bus failover to take advantage of multibus support in
Tru64 UNIX Version 5.0A and multiple-bus failover mode and the ability
to create a NSPOF cluster.
The change in failover modes cannot be accomplished with a simple SET
MULTIBUS COPY=THIS HSG80 CLI command because:
•
Unit offsets are not changed by the HSG80 SET MULTIBUS_FAILOVER
COPY=THIS command.
6–54 Using Fibre Channel Storage
Each path between a Fibre Channel host bus adapter in a host computer
and an active host port on an HSG80 controller is a connection. During
Fibre Channel initialization, when a controller becomes aware of a
connection to a host bus adapter through a switch, it adds the connection
to its table of known connections. The unit offset for the connection
depends on the failover mode in effect at the time the connection is
discovered. In transparent failover mode, host connections to port 1
default to an offset of 0; host connections on port 2 default to an offset
of 100. Host connections on port 1 can see units 0 through 99; host
connections on port 2 can see units 100 through 199.
In multiple-bus failover mode, host connections on either port 1 or 2
can see units 0 through 199. In multiple-bus failover mode, the default
offset for both ports is 0.
If you change the failover mode from transparent failover to multiple-bus
failover, the offsets in the table of known connections remain the same as
if they were for transparent failover mode; the offset on port 2 remains
100. With an offset of 100 on port 2, a host cannot see units 0 through 99
on port 2. This reduces the availability. Also, if you have only a single
HSG80 controller and lose the connection to port 1, you lose access to
units 0 through 99.
Therefore, if you want to change from transparent failover to
multiple-bus failover mode, you must change the offset in the table of
known connections for each connection that has a nonzero offset.
_____________________ Note _____________________
It would do no good to disconnect and then reconnect the
cables, because once a connection is added to the table it
remains in the table until you delete the connection.
•
The system can access a storage device through only one HSG80 port.
The system’s view of the storage device is not changed when the HSG80
is placed in multiple-bus failover mode.
In transparent failover mode, the system accesses storage units D0
through D99 through port 1 and units D100 through D199 through port
2. In multiple-bus failover mode, you want the system to be able to
access all units through all four ports.
To change from transparent failover to multiple-bus failover mode by
resetting the unit offsets and modifying the systems’ view of the storage
units, follow these steps:
1. Shut down the operating systems on all host systems that are accessing
the HSG80 controllers you want to change from transparent failover to
multiple-bus failover mode.
Using Fibre Channel Storage 6–55
2. At the HSG80, set multiple-bus failover as follows. Note that before
putting the controllers in multiple-bus failover mode, you must remove
any previous failover mode:
HSG80> SET NOFAILOVER
HSG80> SET MULTIBUS_FAILOVER COPY=THIS
____________________
Note _____________________
Use the controller known to have the good configuration
information.
3. Execute the SHOW CONNECTION command to determine which
connections have a nonzero offset as follows:
HSG80> SHOW CONNECTION
Connection
Unit
Name
Operating system
Controller Port
THIS
Address
230813
Status
Offset
!NEWCON49
TRU64_UNIX
2
OL this
100
HOST_ID=1000-0000-C920-DA01
ADAPTER_ID=1000-0000-C920-DA01
!NEWCON50
!NEWCON51
!NEWCON52
!NEWCON53
!NEWCON54
!NEWCON55
!NEWCON56
!NEWCON57
!NEWCON58
!NEWCON59
!NEWCON60
!NEWCON61
TRU64_UNIX
HOST_ID=1000-0000-C920-DA01
THIS
1
230813
OL this
0
ADAPTER_ID=1000-0000-C920-DA01
TRU64_UNIX
HOST_ID=1000-0000-C920-EDEB
THIS
2
230913
OL this
100
ADAPTER_ID=1000-0000-C920-EDEB
TRU64_UNIX
HOST_ID=1000-0000-C920-EDEB
THIS
1
230913
OL this
0
0
0
ADAPTER_ID=1000-0000-C920-EDEB
TRU64_UNIX
HOST_ID=1000-0000-C920-EDEB
OTHER
1
230913
OL other
ADAPTER_ID=1000-0000-C920-EDEB
TRU64_UNIX
HOST_ID=1000-0000-C920-DA01
OTHER
1
230813
OL other
ADAPTER_ID=1000-0000-C920-DA01
TRU64_UNIX
HOST_ID=1000-0000-C920-EDEB
OTHER
2
230913
OL other
100
ADAPTER_ID=1000-0000-C920-EDEB
TRU64_UNIX
HOST_ID=1000-0000-C920-DA01
OTHER
2
230813
OL other
100
ADAPTER_ID=1000-0000-C920-DA01
TRU64_UNIX
HOST_ID=1000-0000-C921-09F7
THIS
2
offline
100
ADAPTER_ID=1000-0000-C921-09F7
TRU64_UNIX
HOST_ID=1000-0000-C921-09F7
OTHER
1
offline
0
0
ADAPTER_ID=1000-0000-C921-09F7
TRU64_UNIX
HOST_ID=1000-0000-C921-09F7
THIS
1
offline
ADAPTER_ID=1000-0000-C921-09F7
TRU64_UNIX
HOST_ID=1000-0000-C921-09F7
OTHER
2
offline
100
ADAPTER_ID=1000-0000-C921-09F7
TRU64_UNIX
THIS
2
210513
OL this
100
HOST_ID=1000-0000-C921-086C
ADAPTER_ID=1000-0000-C921-086C
6–56 Using Fibre Channel Storage
!NEWCON62
!NEWCON63
!NEWCON64
!NEWCON65
TRU64_UNIX
HOST_ID=1000-0000-C921-086C
OTHER
1
210513
OL other
0
0
0
ADAPTER_ID=1000-0000-C921-086C
TRU64_UNIX
HOST_ID=1000-0000-C921-0943
OTHER
1
offline
ADAPTER_ID=1000-0000-C921-0943
TRU64_UNIX
HOST_ID=1000-0000-C920-EDA0
OTHER
1
210413
OL other
ADAPTER_ID=1000-0000-C920-EDA0
TRU64_UNIX
OTHER
2
210513
OL other
100
HOST_ID=1000-0000-C921-086C
ADAPTER_ID=1000-0000-C921-086C
.
.
.
The following connections are shown to have nonzero offsets:
!NEWCON49, !NEWCON51, !NEWCON55, !NEWCON56, !NEWCON57,
!NEWCON60, !NEWCON61, and !NEWCON65
4. Set the unit offset to 0 for each connection that has a nonzero unit offset:
HSG80> SET !NEWCON49 UNIT_OFFSET = 0
HSG80> SET !NEWCON51 UNIT_OFFSET = 0
HSG80> SET !NEWCON55 UNIT_OFFSET = 0
HSG80> SET !NEWCON56 UNIT_OFFSET = 0
HSG80> SET !NEWCON57 UNIT_OFFSET = 0
HSG80> SET !NEWCON60 UNIT_OFFSET = 0
HSG80> SET !NEWCON61 UNIT_OFFSET = 0
HSG80> SET !NEWCON65 UNIT_OFFSET = 0
5. At the console of each system accessing storage units on this HSG80,
follow these steps:
a. Use the wwid manager to show the Fibre Channel environment
variables and determine which units are reachable by the system.
This is the information the console uses, when not in wwidmgr
mode, to find Fibre Channel devices:
P00>>> wwidmgr -show ev
wwid0
wwid1
wwid2
wwid3
N1
133 1 WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-002e
131 1 WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-002f
132 1 WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-0030
50001fe100000d64
N2
N3
N4
__________________ Note ___________________
You must set the console to diagnostic mode to use
the wwidmgr command for the following AlphaServer
systems: AS1200, AS4x00, AS8x00, GS60, GS60E, and
GS140. Set the console to diagnostic mode as follows:
P00>>> set mode diag
Console is in diagnostic mode
Using Fibre Channel Storage 6–57
P00>>>
b. For each wwidn line, record the unit number (131, 132, and 133)
and worldwide name for the storage unit. The unit number is the
first field in the display (after wwidn). The Nn value is the HSG80
port being used to access the storage units (controller B, port 2).
c. Clear the wwidn and Nn environment variables:
P00>>> wwidmgr -clear all
d. Initialize the console:
P00>>> init
e. Use the wwid manager with the -quickset option to set up the
device and port path information for the storage units that each
system will need to boot from. Each system may need to boot from
the base operating system disk. Each system will need to boot
from its member system boot disk. Using the storage units from
the example, cluster member 1 will need access to the storage units
with UDIDs 131 (member 1 boot disk) and 133 (Tru64 UNIX disk).
Cluster member 2 will need access to the storage units with UDIDs
132 (member 2 boot disk) and 133 (Tru64 UNIX disk). Set up the
device and port path for cluster member 1 as follows:
P00>>> wwidmgr -quickset -udid 131
.
.
.
P00>>> wwidmgr -quickset -udid 133
.
.
.
f.
Initialize the console:
P00>>> init
g. Verify that the storage units and port path information is set up,
and then reinitialize the console. The following example shows the
information for cluster member 1:
P00>>> wwidmgr -show ev
wwid0
wwid1
wwid2
wwid3
N1
N2
N3
N4
133 1 WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-002e
131 1 WWID:01000010:6000-1fe1-0000-0d60-0009-8080-0434-002f
50001fe100000d64
50001fe100000d62
50001fe100000d63
50001fe100000d61
P00>>> init
h. Set the bootdef_dev console environment variable to the member
system boot device. Use the paths shown in the reachability display
6–58 Using Fibre Channel Storage
of the wwidmgr -quickset command for the appropriate device
(see Section 6.8).
i.
Repeat steps a through h on each system accessing devices on the
HSG80.
6.12 Using the emx Manager to Display Fibre Channel
Adapter Information
The emx manager (emxmgr) utility was written for the TruCluster Software
Product Version 1.6 products to be used to modify and maintain emx driver
worldwide name to target ID mappings. It is included with Tru64 UNIX
Version 5.0A and, although not needed to maintain worldwide name to
target ID mappings, it may be used with TruCluster Server Version 5.0A to:
•
•
•
Display the presence of KGPSA Fibre Channel adapters
Display the target ID mappings for a Fibre Channel adapter
Display the current Fibre Channel topology for a Fibre Channel adapter
See emxmgr(8) for more information on the emxmgr utility.
6.12.1 Using the emxmgr Utility to Display Fibre Channel Adapter
Information
The primary use of the emxmgr utility for TruCluster Server is to display
Fibre Channel information.
Use the emxmgr -d command to display the presence of KGPSA Fibre
Channel adapters on the system. For example:
# /usr/sbin/emxmgr -d
emx0 emx1 emx2
Use the emxmgr -m command to display an adapter ’s target ID mapping.
For example:
# /usr/sbin/emxmgr -m emx0
emx0 SCSI target id assignments:
SCSI tgt id
SCSI tgt id
SCSI tgt id
SCSI tgt id
0 : portname 5000-1FE1-0000-0CB2
nodename 5000-1FE1-0000-0CB0
5 : portname 1000-0000-C920-A7AE
nodename 1000-0000-C920-A7AE
6 : portname 1000-0000-C920-CD9C
nodename 1000-0000-C920-CD9C
7 : portname 1000-0000-C921-0D00
nodename 1000-0000-C921-0D00
(emx0)
Using Fibre Channel Storage 6–59
The previous example shows four Fibre Channel devices on this SCSI bus.
The Fibre Channel adapter in question, emx0, at SCSI ID 7, is denoted by
the presence of the emx0 designation.
Use the emxmgr -t command to display the Fibre Channel topology for
the adapter. For example:
# emxmgr -t emx1
emx1 state information:
Link : connection is UP
Point to Point
1
Fabric attached
FC DID 0x210413
Link is SCSI bus 3 (e.g. scsi3)
SCSI target id 7
portname is 1000-0000-C921-07C4
nodename is 1000-0000-C921-07C4
N_Port at FC DID 0x210013 - SCSI tgt id 5 :
portname 5000-1FE1-0001-8932
2
2
2
2
2
nodename 5000-1FE1-0001-8930
Present, Logged in, FCP Target, FCP Logged in,
N_Port at FC DID 0x210113 - SCSI tgt id 1 :
portname 5000-1FE1-0001-8931
nodename 5000-1FE1-0001-8930
Present, Logged in, FCP Target, FCP Logged in,
N_Port at FC DID 0x210213 - SCSI tgt id 2 :
portname 5000-1FE1-0001-8941
nodename 5000-1FE1-0001-8940
Present, Logged in, FCP Target, FCP Logged in,
N_Port at FC DID 0x210313 - SCSI tgt id 4 :
portname 5000-1FE1-0001-8942
nodename 5000-1FE1-0001-8940
Present, Logged in, FCP Target, FCP Logged in,
N_Port at FC DID 0x210513 - SCSI tgt id 6 :
portname 1000-0000-C921-07F4
nodename 2000-0000-C921-07F4
Present, Logged in, FCP Initiator, FCP Target, FCP Logged in,
N_Port at FC DID 0xfffffc - SCSI tgt id -1 :
portname 20FC-0060-6900-5A1B
3
3
nodename 1000-0060-6900-5A1B
Present, Logged in, Directory Server,
N_Port at FC DID 0xfffffe - SCSI tgt id -1 :
portname 2004-0060-6900-5A1B
nodename 1000-0060-6900-5A1B
Present, Logged in, F_PORT,
1
Status of the emx1 link. The connection is a point-to-point fabric
(switch) connection, and the link is up. The adapter is on SCSI bus 3
at SCSI ID 7. Both the port name and node name of the adapter (the
6–60 Using Fibre Channel Storage
worldwide name) are provided. The Fibre Channel DID number is the
physical Fibre Channel address being used by the N_Port.
2
A list of all other Fibre Channel devices on this SCSI bus, with their
SCSI ID, port name, node name, physical Fibre Channel address and
other items such as:
•
Present — The adapter indicates that this N_Port is present on
the fabric
•
Logged in — The adapter and remote N_Port have exchanged
initialization parameters and have an open channel for
communications (nonprotocol-specific communications)
•
•
FCP Target — This N_Port acts as a SCSI target device (it receives
SCSI commands)
FCP Logged in — The adapter and remote N_Port have exchanged
FCP-specific initialization parameters and have an open channel for
communications (Fibre Channel protocol-specific communications)
•
•
•
Logged Out — The adapter and remote N_Port do not have an open
channel for communication
FCP Initiator — The remote N_Port acts as a SCSI initiator device
(it sends SCSI commands)
FCP Suspended — The driver has invoked a temporary suspension
on SCSI traffic to the N_Port while it resolves a change in
connectivity
•
•
F_PORT — The fabric connection (F_Port) allowing the adapter to
send Fibre Channel traffic into the fabric
Directory Server — The N_Port is the FC entity queried to
determine who is present on the Fibre Channel fabric
3
A target ID of -1 (or -2) shows up for remote Fibre Channel devices
that do not communicate using Fibre Channel protocol, the directory
server, and F_Port.
______________________
Note _______________________
You can use the emxmgr utility interactively to perform any of
the previous functions.
6.12.2 Using the emxmgr Utility Interactively
Start the emxmgr utility without any command-line options to enter the
interactive mode to:
•
Display the presence of KGPSA Fibre Channel adapters
Using Fibre Channel Storage 6–61
•
•
Display the target ID mappings for a Fibre Channel adapter
Display the current Fibre Channel topology for a Fibre Channel adapter
You have already seen how you can perform these functions from the
command line. The same output is available using the interactive mode by
selecting the appropriate option (shown in the following example).
When you start the emxmgr utility with no command-line options, the
default device used is the first Fibre Channel adapter it finds. If you want to
perform functions for another adapter, you must change the targeted adapter
to the correct adapter. For instance, if emx0 is present, when you start the
emxmgr interactively, any commands executed to display information will
provide the information for emx0.
______________________
Note _______________________
The emxmgr has an extensive help facility in the interactive mode.
An example using the emxmgr in the interactive mode follows:
# emxmgr
Now issuing commands to : "emx0"
Select Option (against "emx0"):
1. View adapter’s current Topology
2. View adapter’s Target Id Mappings
3. Change Target ID Mappings
d. Display Attached Adapters
a. Change targeted adapter
x. Exit
----> 2
emx0 SCSI target id assignments:
SCSI tgt id
SCSI tgt id
SCSI tgt id
SCSI tgt id
0 : portname 5000-1FE1-0000-0CB2
nodename 5000-1FE1-0000-0CB0
5 : portname 1000-0000-C920-A7AE
nodename 1000-0000-C920-A7AE
6 : portname 1000-0000-C920-CD9C
nodename 1000-0000-C920-CD9C
7 : portname 1000-0000-C921-0D00
nodename 1000-0000-C921-0D00
(emx0)
Select Option (against "emx0"):
1. View adapter’s current Topology
6–62 Using Fibre Channel Storage
2. View adapter’s Target Id Mappings
3. Change Target ID Mappings
d. Display Attached Adapters
a. Change targeted adapter
x. Exit
----> x
#
Using Fibre Channel Storage 6–63
7
Preparing ATM Adapters
The Compaq Tru64 UNIX operating system supports Asynchronous Transfer
Mode (ATM). TruCluster Server supports the use of LAN emulation over
ATM for client access.
This chapter provides an ATM overview, an example TruCluster Server
cluster using ATM, an ATM adapter installation procedure, and information
about verifying proper installation of fiber optic cables. See the Tru64 UNIX
Asynchronous Transfer Mode manual for information on configuring the
ATM software.
7.1 ATM Overview
In synchronous transfer methods, time-division multiplexing (TDM)
techniques are used to divide the bandwidth into fixed-size channels
dedicated to particular connections. If a system has nothing to transmit
when its time slot comes up, that time slot is wasted. Also, if the system has
lots of information to transmit, the system can only transmit when its turn
comes up, even if other time slots are empty.
ATM eliminates the inefficiencies of TDM technology by sharing network
bandwidth among multiple logical connections. Instead of dividing the
bandwidth into fixed-size channels dedicated to particular connections, ATM
uses the entire bandwidth to transmit a steady stream of fixed-size (53-byte)
cells. Each cell includes a 5-byte header containing an address to identify
the cell with a particular logical connection.
If a connection needs more bandwidth, it is allocated more cells. When a
connection is idle, it uses no cells and consumes no bandwidth. This feature
makes ATM the ideal technology for transferring voice, video, and data
through private networks and across public networks.
ATM is a connection-oriented, cell-switching and multiplexing technology.
Cells transit ATM networks by passing through ATM switches, which
analyze information in the header to switch the cell to the output interface
that connects the cell to the next appropriate switch as the cell proceeds to
its destination.
The ATM switch acts as a hub in the ATM network. All devices are attached
to an ATM switch, either directly or indirectly.
Preparing ATM Adapters 7–1
Most data traffic in existing customer networks is sent over Local Area
Networks (LANs) such as Ethernet or Token Ring networks. The services
provided by the LANs differ from those of ATM, for example:
•
•
•
LAN messages are connectionless; ATM is a connection-oriented
technology
Because a LAN is based on a shared medium, it is easy to broadcast
messages
LAN addresses are based on hardware manufacturing serial numbers
and are independent of the network topology
In order to use the large base of existing LAN application software, ATM
defines a LAN Emulation (LANE) service that emulates services of existing
LANs across an ATM network.
The LAN emulation environment groups hosts into an emulated LAN
(ELAN) which has the following characteristics:
•
•
Identifies hosts through their 48-bit media access control (MAC) number
Supports multicast and broadcast services through point-to-multipoint
connections or through a multicast server
•
•
Supports any protocol that uses an IEEE broadcast LAN
Provides the appearance of a connectionless service to participating end
systems
One or more emulated LANs can run on the same ATM network. Each ELAN
is independent of the others and users cannot communicate directly across
emulated LAN boundries. Communication between ELANs is possible only
through routers or bridges.
Each ELAN is composed of:
•
A set of LAN emulation clients (LECs): An LEC resides in each end
system and performs data forwarding, address resolution, and control
functions that provide a MAC-level emulated Ethernet interface to
higher-level software and other entities within the emulated LAN.
•
A LAN emulation service, which normally resides on an ATM switch
and consists of:
–
LAN Emulation Configuration Server (LECS): An LECS implements
the assignment of individual LAN emulation clients to different
emulated LANs. It provides the client with the ATM address of the
LAN emulation server.
–
LAN Emulation Server (LES): An LES implements the control
coordination function for the emulated LAN by registering and
resolving MAC addresses and route descriptors to ATM addresses.
7–2 Preparing ATM Adapters
–
Broadcast and Unknown Server (BUS): A BUS handles broadcast
data sent by a LAN emulation client, all multicast data, and data sent
by a LAN emulation client before the ATM address has been resolved.
Figure 7–1 shows an ATM network with two emulated LANs. Hosts A and B
are LECs on ELAN1. Hosts C, D, and E are LECs on ELAN2. The LECS, the
LES, and the BUS are server functions resident on the ATM switch (even
though they are shown separately).
Figure 7–1: Emulated LAN Over an ATM Network
LECS
LES
LECS
LES
Host D
BUS
BUS
ELAN 2
ATM Switch
ATM Switch
Host A
ELAN 1
Host C
Host E
Host B
ZK-1323U-AI
Use LAN emulation over ATM in a TruCluster Server cluster for client
system access (Memory Channel is the cluster interconnect).
7.2 Installing ATM Adapters
_____________________ Warning _____________________
Some fiber optic equipment can emit laser light that can injure
your eyes. Never look into an optical fiber or connector port.
Always assume the cable is connected to a light source.
______________________
Note _______________________
Do not touch the ends of the fiber optics cable. The oils from your
skin can cause an optical power loss.
Preparing ATM Adapters 7–3
Use the following steps to install an ATMworks adapter. See the ATMworks
350 Adapter Installation and Service guide for more information. Be sure
to use the antistatic ground strap.
1. Remove the adapter extender bracket if the ATMworks 350 is to be
installed in an AlphaServer 2100 system.
2. Remove the option slot cover from the appropriate PCI or
TURBOchannel slot.
3. Install the adapter module.
4. Install the multimode fiber optics (SC connectors) cables as follows:
•
•
Remove the optical dust caps.
Line up the transmit cable connector with the transmit port and
the receive cable connector with the receive port and insert the SC
connectors. The ATMworks transmit port is identified by an arrow
exiting a circle. The receive port is identified by an arrow entering a
circle.
Listen for the click indicating that the connector is properly seated.
___________________ Note
___________________
Ensure that the bend radius of any fiber optic cable
exceeds 2.5 cm (1 inch) to prevent breaking the glass.
When removing an SC connector, do not pull on the cable.
Pull on the cable connector only.
To verify that the cables are connected correctly, see Section 7.3.
7.3 Verifying ATM Fiber Optic Cable Connectivity
The fiber optic cables from some suppliers are not labeled or color coded,
and as the system and ATM switch may be separated by a great distance,
verifying that the cables are connected correctly may be difficult.
The ATMworks adapters start sending idle cells when the ATM driver is
enabled. The adapter sends idle cells even when no data is being sent. ATM
switches provide an indication that they are receiving the idle cells.
To verify that the fiber optic cables are properly connected, follow these steps:
1. Verify that both the transmit and receive connectors are seated properly
at both the ATM adapter and the ATM switch.
2. Verify that the following ATM subsets have been installed with this
command:
7–4 Preparing ATM Adapters
# /usr/sbin/setld -i | grep ATM
•
•
OSFATMBASE: ATM Commands
OSFATMBIN: ATM Kernel Modules
Additionally, after the ATM subsets have been installed, verify that a
new kernel has been built with the following kernel options selected
(/sbin/sysconfig -q atm):
•
•
•
Asynchronous Transfer Mode (ATM)
ATM UNI 3.0/3.1 Signalling for SVCs
LAN Emulation over ATM (LANE)
3. Enable the ATM driver with the following command:
# /usr/sbin/atmconfig up driver=driver_name
In the command, driver_name is lta# for the ATMworks 350. The
number sign (#) is the adapter number.
To enable lta0 to initiate contact with the network, enter the following
command:
# /usr/sbin/atmconfig up driver=lta0
4. Check the ATM switch for an indication that it is receiving idle cells.
The following table provides the indication for a few ATM switches. If
you do not have one of these switches, check the documentation for your
switch to determine how the switch indicates that it is cabled correctly.
Indicator
ATM Switch
Comments
Compaq GIGAswitch
PHY
Illuminated green LED indicates
that the switch is receiving idle
cells from the ATM adapter.
Bay Networks®
Centillion100
En
Illuminated green LED indicates
that the switch is receiving idle
cells from the ATM adapter.
SynOptics® LattisCell
10114
Link
TX
Illuminated green LED indicates
that the switch is receiving idle
cells from the ATM adapter.
CISCO® Systems
LightStream 1010
The switch starts transmitting
data as soon as it receives idle
cells. The green TX LED will
flash on and off.
FORESystems®
ForeRunner A S X200
TX
The Yellow TX LED will
be on steady.
Preparing ATM Adapters 7–5
5. If you do not have an indication that confirms a correct cable connection,
swap the transmit and receive connectors on one end of the cable and
recheck the indicators.
6. If you still do not have a correct cable connection, you probably have a
bad cable.
7.4 ATMworks Adapter LEDs
The ATMworks adapter has two LEDS that indicate the status of the
adapter and its connections to the network, the Network LED, and the
Module LED. The Network LED is labeled with a number sign (#) under the
LED. The Module LED is labeled with an incomplete circle under the LED.
The meaning of the LEDs is shown in Table 7–1.
Table 7–1: ATMworks Adapter LEDs
Network LED
Module LED
Description
Off
Off
PCI slot is not receiving power, or the
ATMworks driver has not been loaded.
Off/Amber/Green
Green
ATMworks driver is loaded and
the module is OK.
Amber
Off
Amber
ATMworks adapter is in reset mode.
The adapter diagnostics failed.
Amber
Green
Amber
Green/Off
Green
A physical link connection has been made.
There is no physical link connection.
7–6 Preparing ATM Adapters
8
Configuring a Shared SCSI Bus for Tape
Drive Use
The topics in this section provide information on preparing the various tape
devices for use on a shared SCSI bus with the TruCluster Server product.
______________________ Notes ______________________
Section 8.6 and Section 8.7 provide documentation for
the TL890/TL891/TL892 MiniLibrary family as sold with
the DS-TL891-NE/NG, DS-TL891-NT, DS-TL892-UA,
DS-TL890-NE/NG part numbers.
The TL881, with a Compaq 6-3 part number was recently
qualified in cluster configurations. The TL891 rackmount base
unit has been provided with a Compaq 6-3 part number. The
TL881 and TL891 only differ in the type of tape drive they use.
They both work with an expansion unit (previously called the
DS-TL890-NE) and a new module called the data unit.
Section 8.11 covers the TL881 and TL891 with the common
components as sold with the Compaq 6-3 part numbers.
As long as the TL89x MiniLibrary family is being sold with
both sets of part numbers, the documentation will retain the
documentation for both ways to configure the MiniLibrary.
8.1 Preparing the TZ88 for Shared Bus Usage
Two versions of the TZ88 are supported, the TZ88N-TA tabletop standalone
enclosure, and the TZ88N-VA StorageWorks building blocks (SBB) 5.25-inch
carrier.
As with any of the shared SCSI devices, the TZ88N-TA and TZ88N-VA SCSI
IDs must be set to ensure that no two SCSI devices on the shared SCSI
bus have the same SCSI ID.
The following sections describe preparing the TZ88 in more detail.
Configuring a Shared SCSI Bus for Tape Drive Use 8–1
8.1.1 Setting the TZ88N-VA SCSI ID
You must set the TZ88N-VA switches before the tape drive is installed into
the BA350 StorageWorks enclosure. The Automatic selection is normally
used. The TZ88N-VA takes up three backplane slot positions. The physical
connection is in the lower of the three slots. For example, if the tape drive is
installed in slots 1, 2, and 3 with the switches in Automatic, the SCSI ID
is 3. If the tape drive is installed in slots 3, 4, and 5 with the switches in
Automatic, the SCSI ID is 5. The switch settings are shown in Table 8–1.
Figure 8–1 shows the TZ88N-VA with the backplane interface connector and
SCSI ID switch pack.
Figure 8–1: TZ88N-VA SCSI ID Switches
Backplane
Interface
Connector
SCSI ID
Switch Pack
Snap−in
Locking
Handles
TZ88N−VA
8–2 Configuring a Shared SCSI Bus for Tape Drive Use
Table 8–1: TZ88N-VA Switch Settings
SCSI ID
SCSI ID Selection Switches
1
2
3
4
5
6
Automatica
Off
Off
On
Off
On
Off
On
Off
On
Off
Off
Off
On
On
Off
Off
On
On
Off
Off
Off
Off
Off
On
On
On
On
On
Off
Off
Off
Off
Off
Off
Off
Off
On
Off
Off
Off
Off
Off
Off
Off
Off
On
Off
Off
Off
Off
Off
Off
Off
Off
0
1
2
3
4
5
6
7
a
SBB tape drive SCSI ID is determined by the SBB physical slot.
8.1.2 Cabling the TZ88N-VA
There are no special cabling restrictions specific to the TZ88N-VA; it is
installed in a BA350 StorageWorks enclosure. A DWZZA-VA installed in slot
0 of the BA350 provides the connection to the shared SCSI bus. The tape
drive takes up three slots, so two SCSI IDs are unavailable for disks in this
StorageWorks enclosure. Another BA350 may be daisy chained to allow the
use of the SCSI IDs unavailable in the first StorageWorks enclosure due to
the TZ88 tape drive.
You must remove the DWZZA-VA differential terminators. Ensure that
DWZZA-VA jumper J 2 is installed to enable the single-ended termination.
The BA350 jumper and terminator must be installed.
A trilink connector on the DWZZA-VA differential end allows connection to
the shared bus. An H879-AA terminator is installed on the trilink for the
BA350 on the end of the bus to provide shared SCSI bus termination.
Figure 8–2 shows a TruCluster Server cluster with two shared SCSI buses.
The top shared bus has a BA350 with disks at SCSI IDs 1, 2, 4, and 5. The
other BA350 contains a TZ88N-VA at SCSI ID 3.
Configuring a Shared SCSI Bus for Tape Drive Use 8–3
Figure 8–2: Shared SCSI Buses with SBB Tape Drives
KZPSA adapter,
trilink connector,
and H879 terminator
BN21K or BN21L cable
DWZZA-VA
and trilink
BN21K or BN21L cables
DWZZA-VA,
trilink,
and H879
terminator
T
T
T
1
2
TZ88N-VA
4
5
Memory
Channel
adapters
Memory Channel link cable
BA350
BA350
AlphaServer 2100A
AlphaServer 2100A
DWZZB-VW, trilink connector,
and H879 terminator
DWZZB-VW and trilink
T
1
TZ89N-VW
3
4
5
BA356
BA356
ZK-1334U-AI
8.1.3 Setting the TZ88N-TA SCSI ID
The TZ88N-TA SCSI ID is set with a push-button counter switch on the rear
of the unit. Push the button above the counter to increment the address;
push the button below the counter to decrement the address until you have
the desired SCSI ID selected.
8.1.4 Cabling the TZ88N-TA
You must connect the TZ88N-TA tabletop model to a single-ended segment
of the shared SCSI bus. It is connected to a differential portion of the
shared SCSI bus with a DWZZA-AA or DWZZB-AA. Figure 8–6 shows a
configuration of a TZ885 for use on a shared SCSI bus. You can replace the
TZ885 shown in the illustration with a TZ88N-TA. To configure the shared
SCSI bus for use with a TZ88N-TA, follow these steps:
1. You will need one DWZZA-AA or DWZZB-AA for each TZ88N-TA.
8–4 Configuring a Shared SCSI Bus for Tape Drive Use
Ensure that DWZZA jumper J 2 or DWZZB jumpers W1 and W2 are
installed to enable the single-ended termination.
Remove the termination from the differential end by removing the five
14-pin SIP resistors.
2. Attach a trilink connector to the differential end of the DWZZA or
DWZZB.
3. Connect the single-ended end of a DWZZA to the TZ88N-TA with a
BC19J cable.
Connect the single-ended end of a DWZZB to the TZ88N-TA with a
BN21M cable.
4. Install a H8574-A or H8890-AA terminator on the other TZ88N-TA
SCSI connector.
5. Connect a trilink or Y cable to the differential shared SCSI bus with
BN21K or BN21L cables. Ensure that the trilink or Y cable at the end
of the bus is terminated with an H879-AA terminator.
The single-ended SCSI bus may be daisy chained from one single-ended
tape drive to another with BC19J cables as long as the SCSI bus maximum
length is not exceeded. Ensure that the tape drive on the end of the bus is
terminated with a H8574-A or H8890-AA terminator.
You can add additional TZ88N-TA tape drives to the differential shared SCSI
bus by adding additional DWZZA or DWZZB/TZ88N-TA combinations.
______________________
Note _______________________
Ensure that there is no conflict with tape drive, system, and disk
SCSI IDs.
8.2 Preparing the TZ89 for Shared SCSI Usage
Like the TZ88, the TZ89 comes in either a tabletop (DS-TZ89N-TA) or a
StorageWorks building block (SBB) 5.25-inch carrier (DS-TZ89N-VW). The
SBB version takes up three slots in a BA356 StorageWorks enclosure.
The following sections describe how to prepare the TZ89 in more detail.
8.2.1 Setting the DS-TZ89N-VW SCSI ID
The DS-TZ89N-VW backplane connector makes a connection with the
backplane in the middle of the three slots occupied by the drive. If the
switches are set to automatic to allow the backplane position to select the
SCSI ID, the ID corresponds to the backplane position of the middle slot.
For example, if the DS-TZ89N-VW is installed in a BA356 in slots 1, 2, and
Configuring a Shared SCSI Bus for Tape Drive Use 8–5
3, the SCSI ID is 2. If it is installed in slots 3, 4, and 5, the SCSI ID is
4. Figure 8–3 shows a view of the DS-TZ89N-VW showing the backplane
interface connector and SCSI ID switch pack.
Figure 8–3: DS-TZ89N-VW SCSI ID Switches
Backplane
Interface
Connector
SCSI ID
Switch Pack
Snap−in
Locking
Handles
DS−TZ89N−VW
The SCSI ID is selected by switch positions, which must be selected before
the tape drive is installed in the BA356. Table 8–2 shows the switch settings
for the DS-TZ89N-VW.
Table 8–2: DS-TZ89N-VW Switch Settings
SCSI ID
SCSI ID Selection Switches
1
2
3
4
5
6
7
8
Automatica
Off
Off
On
Off
Off
Off
Off
Off
Off
Off
Off
Off
On
Off
Off
On
Off
Off
On
Off
Off
On
Off
Off
0
1
8–6 Configuring a Shared SCSI Bus for Tape Drive Use
Table 8–2: DS-TZ89N-VW Switch Settings (cont.)
SCSI ID
SCSI ID Selection Switches
2
Off
On
Off
On
Off
On
Off
On
Off
On
Off
On
Off
On
On
On
Off
Off
On
On
Off
Off
On
On
Off
Off
On
On
Off
Off
On
On
On
On
Off
Off
Off
Off
On
On
On
On
Off
Off
Off
Off
Off
Off
On
On
On
On
On
On
On
On
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
Off
3
4
5
6
7
8
9
10
11
12
13
14
15
a
SBB tape drive SCSI ID is determined by the SBB physical slot.
8.2.2 Cabling the DS-TZ89N-VW Tape Drives
No special cabling is involved with the DS-TZ89N-VW as it is installed in
a BA356 StorageWorks enclosure. A DWZZB-VA installed in slot 0 of the
BA356 provides the connection to the shared SCSI bus.
You must remove the DWZZB-VW differential terminators. Ensure that
jumpers W1 and W2 are installed to enable the single-ended termination.
The BA356 jumper must be installed, and connector J B1 on the personality
module must be left open to provide termination at the other end of the
single-ended bus.
A trilink connector on the differential end of the DWZZB-VW allows
connection to the shared bus. If the BA356 containing the DS-TZ89N-VW is
on the end of the bus, install an H879-AA terminator on the trilink for that
BA356 to provide termination for the shared SCSI bus.
Figure 8–2 shows a TruCluster Server cluster with two shared SCSI buses.
The bottom shared bus has a BA356 with disks at SCSI IDs 1, 3, 4, and 5.
The other BA356 contains a DS-TZ89N-VW at SCSI ID 2.
Configuring a Shared SCSI Bus for Tape Drive Use 8–7
8.2.3 Setting the DS-TZ89N-TA SCSI ID
The DS-TZ89N-TA has a push-button counter switch on the rear panel to
select the SCSI ID. It is preset at the factory to 15. Push the button above
the counter to increment the SCSI ID (the maximum is 15); push the button
below the switch to decrease the SCSI ID.
8.2.4 Cabling the DS-TZ89N-TA Tape Drives
You must connect the DS-TZ89N-TA tabletop model to a single-ended
segment of the shared SCSI bus. It is connected to a differential portion of
the shared SCSI bus with a DWZZB-AA. Figure 8–6 shows a configuration of
a T885 for use on a shared SCSI bus. J ust replace the TZ885 in the figure
with a DS-TZ89N-TA and the DWZZA-AA with a DWZZB-AA. To configure
the shared SCSI bus for use with a DS-TZ89N-TA follow these steps:
1. You will need one DWZZB-AA for each DS-TZ89N-TA.
Ensure that the DWZZB jumpers W1 and W2 are installed to enable the
single-ended termination.
Remove the termination from the differential end by removing the five
14-pin SIP resistors.
2. Attach a trilink connector to the differential end of the DWZZB-AA.
3. Connect the DWZZB-AA single-ended end to the DS-TZ89N-TA with a
BN21K or BN21L cable.
4. Install an H879-AA terminator on the other DS-TZ89N-TA SCSI
connector.
5. Connect the trilink to the differential shared SCSI bus with BN21K
or BN21L cables. Ensure that the trilink at the end of the bus is
terminated with an H879-AA terminator.
The wide, single-ended SCSI bus may be daisy chained from one single-ended
tape drive to another with BN21K or BN21L cables as long as the SCSI bus
maximum length is not exceeded. Ensure that the tape drive on the end of
the bus is terminated with an H879-AA terminator.
You can add additional DS-TZ89N-TA tape drives to the differential shared
SCSI bus by adding additional DWZZB-AA/DS-TZ89N-TA combinations.
______________________
Note _______________________
Ensure that there is no conflict with tape drive, system, and disk
SCSI IDs.
8–8 Configuring a Shared SCSI Bus for Tape Drive Use
8.3 Compaq 20/40 GB DLT Tape Drive
The Compaq 20/40 GB DLT Tape Drive is a Digital Linear Tape (DLT)
tabletop cartridge tape drive capable of holding up to 40 GB of data
per Compactape IV cartridge using 2:1 compression. It is capable of
storing/retrieving data at a rate of up to 10.8 GB per hour (using 2:1
compression).
The Compaq 20/40 GB DLT Tape Drive uses CompacTape III, CompacTape
IIIXT, or CompacTape IV media.
It is a narrow, single-ended SCSI device, and uses 50-pin, high-density
connectors.
For more information on the Compaq 20/40 GB DLT Tape Drive, see the
following Compaq documentation:
•
•
Compaq DLT User Guide (185292-002)
DLT Tape Drive User Guide Supplement (340949-002)
The following sections describe how to prepare the Compaq 20/40 GB DLT
Tape Drive for shared SCSI bus usage in more detail.
8.3.1 Setting the Compaq 20/40 GB DLT Tape Drive SCSI ID
As with any of the shared SCSI devices, the Compaq 20/40 GB DLT Tape
Drive SCSI ID must be set to ensure that no two SCSI devices on the shared
SCSI bus have the same SCSI ID.
The Compaq 20/40 GB DLT Tape Drive SCSI ID is set with a push-button
counter switch on the rear of the unit (see Figure 8–4). Push the button above
the counter to increment the address; push the button below the counter to
decrement the address until you have the desired SCSI ID selected.
Only SCSI IDs in the range of 0 to 7 are valid. Ensure that the tape drive
SCSI ID does not conflict with the SCSI ID of the host bus adapters (usually
6 and 7) or other devices on this shared SCSI bus.
Configuring a Shared SCSI Bus for Tape Drive Use 8–9
Figure 8–4: Compaq 20/40 GB DLT Tape Drive Rear Panel
SCSI ID
SCSI ID
Selector
Switch
+
+
0
-
0
-
20/40 GB DLT Tape Drive
ZK-1603U-AI
8.3.2 Cabling the Compaq 20/40 GB DLT Tape Drive
The Compaq 20/40 GB DLT Tape Drive is connected to a single-ended
segment of the shared SCSI bus. A DWZZB-AA signal converter is required
to convert the differential shared SCSI bus to single-ended. Figure 8–5
shows a configuration with a Compaq 20/40 GB DLT Tape Drive on a shared
SCSI bus.
To configure the shared SCSI bus for use with a Compaq 20/40 GB DLT
Tape Drive, follow these steps:
1. You will need one DWZZB-AA for each shared SCSI bus with a Compaq
20/40 GB DLT Tape Drive.
Ensure that the DWZZB-AA jumpers W1 and W2 are installed to enable
the single-ended termination.
Remove the termination from the differential end by removing the five
14-pin SIP resistors.
2. Attach an H885-AA trilink connector or BN21W-0B Y cable to the
differential end of the DWZZB-AA.
3. Connect the single-ended end of the DWZZB-AA to the Compaq 20/40
GB DLT Tape Drive with cable part number 199629-002 or 189636-002
(1.8-meter cables).
4. Install terminator part number 341102-001 on the other tape drive
SCSI connector.
5. Connect the trilink on the DWZZB-AA to another trilink or Y cable
on the differential shared SCSI bus with a 328215-00X, BN21K, or
BN21L cable. Keep the length of the differential segment below the
25-meter maximum length (cable part number 328215-004 is a 20-meter
8–10 Configuring a Shared SCSI Bus for Tape Drive Use
cable). Ensure that the trilink or Y cable at both ends of the differential
segment of the shared SCSI bus is terminated with an HD68 differential
terminator such as an H879-AA.
The single-ended SCSI bus may be daisy chained from one single-ended
tape drive to another with cable part number 146745-003 or 146776-003
(0.9-meter cables) as long as the SCSI bus maximum length of 3 meters (fast
SCSI) is not exceeded. Ensure that the tape drive on the end of the bus is
terminated with terminator part number 341102-001.
You can add additional shared SCSI buses with Compaq 20/40 GB DLT
Tape Drives by adding additional DWZZB-AA/Compaq 20/40 GB DLT Tape
Drive combinations.
______________________ Notes ______________________
Ensure that there is no conflict with tape drive and host bus
adapter SCSI IDs.
To achieve system performance capabilities, we recommend
placing no more than two Compaq 20/40 GB DLT Tape Drives on
a SCSI bus, and also recommend that no shared storage be placed
on the same SCSI bus with the tape drive.
Configuring a Shared SCSI Bus for Tape Drive Use 8–11
Figure 8–5: Cabling a Shared SCSI Bus with a Compaq 20/40 GB DLT Tape
Drive
Network
Member
System
1
Member
System
2
Memory
Channel
Interface
7
Memory Channel
KZPBA-CB (ID 6)
Memory Channel
KZPBA-CB (ID 7)
KZPBA-CB (ID 7)
7
T
6
5
5
T
KZPBA-CB (ID 6)
T
1
1
+
0
-
T
DS-DWZZH-03
T
T
T
T
6
8
10
9
DWZZB-AA
2
3
20/40 GB DLT
Tape Drive
4
T
Controller B
HSZ70
Controller A
HSZ70
StorageWorks
RAID Array 7000
NOTE: This drawing is not to scale.
ZK-1604U-AI
Table 8–3 shows the components used to create the cluster shown in
Figure 8–5.
Table 8–3: Hardware Components Used to Create the Configuration Shown
in Figure 8 — 5
Callout Number
Description
BN38C or BN38D cablea
BN37A cableb
1
2
3
4
5
6
7
8
H8861-AA VHDCI trilink connector
H8863-AA VHDCI terminator
BN21W-0B Y cable
H879-AA terminator
328215-00X, BN21K, or BN21L cablec
H885-AA trilink connector
8–12 Configuring a Shared SCSI Bus for Tape Drive Use
Table 8–3: Hardware Components Used to Create the Configuration Shown
in Figure 8 — 5 (cont.)
Description
Callout Number
9
199629-002 or 189636-002 (1.8-meter cable)
341102-001 terminator
10
a
b
c
The maximum length of the BN38C (or BN38D) cable on one SCSI bus segment must not exceed 25 meters.
The maximum length of the BN37A cable must not exceed 25 meters.
The maximum combined length of these cables must not exceed 25 meters.
8.4 Preparing the TZ885 for Shared SCSI Usage
The TZ885 Digital Linear Tape subsystems combine a cartridge tape drive
(TZ88) and an automatic cartridge loader. The TZ885 uses a removable
magazine.
The TZ885 uses a five-cartridge (CompacTape IV) minitape library
(magazine) with a 200-GB capacity (compressed). It is capable of
reading/writing at approximately 10.8 GB per hour.
As with any of the shared SCSI devices, the TZ885 SCSI IDs must be set
to ensure that no two SCSI devices on the shared SCSI bus have the same
SCSI ID.
The following sections describe preparing the TZ885 in more detail.
8.4.1 Setting the TZ885 SCSI ID
To set the TZ885 SCSI ID from the Operators Control Panel (OCP), follow
these steps:
1. Press and hold the Display Mode push button (for about five seconds)
until the SCSI ID SEL message is displayed:
SCSI ID SEL
SCSI ID 0
2. Press the Select push button until you see the desired SCSI ID number
in the display.
3. Press the Display Mode push button again.
4. Issue a bus reset or turn the minilibrary power off and on again to cause
the drive to recognize the new SCSI ID.
8.4.2 Cabling the TZ885 Tape Drive
The TZ885 is connected to a single-ended segment of the shared SCSI
bus. It is connected to a differential portion of the shared SCSI bus with a
DWZZA-AA or DWZZB-AA. Figure 8–6 shows a configuration of a TZ885
Configuring a Shared SCSI Bus for Tape Drive Use 8–13
for use on a shared SCSI bus. The TZ885 in this figure has had the SCSI
ID set to 0 (zero). To configure the shared SCSI bus for use with a TZ885,
follow these steps:
1. You will need one DWZZA-AA or DWZZB-AA for each TZ885 tape drive.
Ensure that the DWZZA jumper J 2 or DWZZB jumpers W1 and W2 are
installed to enable the single-ended termination.
Remove the termination from the differential end by removing the five
14-pin SIP resistors.
2. Attach a trilink connector to the differential end of the DWZZA or
DWZZB.
3. Connect the single-ended end of a DWZZA to the TZ885 with a BC19J
cable.
Connect the single-ended end of a DWZZB to the TZ885 with a BN21M
cable.
4. Install an H8574-A or H8890-AA terminator on the other TZ885 SCSI
connector.
5. Connect a trilink or Y cable to the differential shared SCSI bus with
BN21K or BN21L cables. Ensure that the trilink or Y cable at the end
of the bus is terminated with an H879-AA terminator.
The single-ended SCSI bus may be daisy chained from one single-ended
tape drive to another with BC19J cables as long as the SCSI bus maximum
length is not exceeded. Ensure that the tape drive on the end of the bus is
terminated with a H8574-A or H8890-AA terminator.
You can add additional TZ885 tape drives to the differential shared SCSI
bus by adding additional DWZZA or DWZZB/TZ885 combinations.
______________________
Note _______________________
Ensure that there is no conflict with tape drive, system, and disk
SCSI IDs.
8–14 Configuring a Shared SCSI Bus for Tape Drive Use
Figure 8–6: Cabling a Shared SCSI Bus with a TZ885
KZPSA adapter,
trilink connector,
and H879 terminator
DWZZA-VA,
trilink connector,
and H879 terminator
DWZZA-VA
and trilink connector
BN21K or BN21L cables
T
T
T
1
1
2
3
2
3
4
5
4
5
Memory
Channel
adapters
BN21K
or
BN21L
cable
Memory Channel link cable
BA350
BA350
AlphaServer 2100A
AlphaServer 2100A
BC19J
T
T
Trilink
connector
and
DWZZA-AA
H879-AA
terminator
TZ885
H8574-A
terminator
ZK-1344U-AI
8.5 Preparing the TZ887 for Shared SCSI Bus Usage
The TZ887 Digital Linear Tape (DLT) MiniLibrary combines a cartridge tape
drive (TZ88) and an automatic cartridge loader. It uses a seven-cartridge
(CompacTape IV) removable magazine with a total capacity of nearly 280
GB compressed. It is capable of reading/writing at approximately 10.8 GB
per hour.
As with any of the shared SCSI devices, the TZ887 SCSI IDs must be set
to ensure that no two SCSI devices on the shared SCSI bus have the same
SCSI ID.
The following sections describe how to prepare the TZ887 in more detail.
8.5.1 Setting the TZ887 SCSI ID
The TZ887 SCSI ID is set with a push-button counter switch on the rear of
the unit (see Figure 8–7). Push the button above the counter to increment
the address; push the button below the counter to decrement the address
until you have the desired SCSI ID selected.
Configuring a Shared SCSI Bus for Tape Drive Use 8–15
Figure 8–7: TZ887 DLT MiniLibrary Rear Panel
SCSI ID
Selector
Switch
SCSI ID
+
+
0
-
0
-
TZ887
ZK-1461U-AI
8.5.2 Cabling the TZ887 Tape Drive
The TZ887 is connected to a single-ended segment of the shared SCSI
bus. It is connected to a differential portion of the shared SCSI bus with a
DWZZB-AA. Figure 8–8 shows a configuration with a TZ887 for use on a
shared SCSI bus. The TZ887 in this figure would have the SCSI ID set to
0. The member systems use SCSI IDs 6 and 7, and the disks are located
in the BA356 slots at SCSI IDs 1-5.
To configure the shared SCSI bus for use with a TZ887, follow these steps:
1. You will need one DWZZB-AA for each shared SCSI bus with a TZ887
tape drive.
Ensure that the DWZZB-AA jumpers W1 and W2 are installed to enable
the single-ended termination.
Remove the termination from the differential end by removing the five
14-pin SIP resistors.
2. Attach an H885-AA trilink connector to the differential end of the
DWZZB-AA.
3. Connect the single-ended end of the DWZZB-AA to the TZ887 with a
BN21M cable.
4. Install an H8574-A or H8890-AA terminator on the other TZ887 SCSI
connector.
5. Connect the trilink on the DWZZB-AA to another trilink or Y cable
on the differential shared SCSI bus with BN21K or BN21L cables.
Ensure that the trilink or Y cable at both ends of the shared SCSI bus is
terminated with an H879-AA terminator.
8–16 Configuring a Shared SCSI Bus for Tape Drive Use
The single-ended SCSI bus may be daisy chained from one single-ended
tape drive to another with BC19J cables, as long as the SCSI bus maximum
length is not exceeded and there are sufficient SCSI IDs available. Ensure
that the tape drive on the end of the bus is terminated with an H8574-A or
H8890-AA terminator.
You can add additional shared SCSI buses with TZ887 tape drives by adding
additional DWZZB-AA/TZ887 combinations.
______________________
Note _______________________
Ensure that there is no conflict with tape drive, host bus adapter,
and disk SCSI IDs.
Figure 8–8: Cabling a Shared SCSI Bus with a TZ887
KZPSA adapter,
trilink connector,
DWZZB-VW,
trilink connector,
and H879-AA terminator
and H879 terminator
BN21K or BN21L cables
DWZZB-VW
and trilink connector
T
T
T
1
1
2
3
2
3
4
5
4
5
BN21K
or
BN21L
cable
Memory
Channel
adapters
Memory Channel link cable
BA356
BA356
AlphaServer 2100A
AlphaServer 2100A
H8574-A
terminator
BN21M
+
0
-
T
T
Trilink
connector
and
DWZZB-AA
H879-AA
terminator
TZ887
ZK-1462U-AI
Configuring a Shared SCSI Bus for Tape Drive Use 8–17
8.6 Preparing the TL891 and TL892 DLT MiniLibraries for
Shared SCSI Usage
______________________
Note _______________________
To achieve system performance capabilities, we recommend
placing no more than two TZ89 drives on a SCSI bus, and also
recommend that no shared storage be placed on the same SCSI
bus with a tape library.
The TL891 and TL892 MiniLibraries use one (TL891) or two (TL892)
TZ89N-AV differential tape drives and a robotics controller, which access
cartridges in a 10-cartridge magazine.
Each tape drive present, and the robotics controller, have individual SCSI
IDs.
There are six 68-pin, high-density SCSI connectors located on the back of
the MiniLibrary; two SCSI connectors for each drive and two for the robotics
controller. The TL891 uses a 0.3-meter SCSI bus jumper cable (part of the
TL891 package) to place the robotics controller and tape drive on the same
SCSI bus. When upgrading to the TL892, you can place the second drive on
the same SCSI bus (another 0.3-meter SCSI bus jumper cable is supplied
with the DS-TL892-UA upgrade kit) or place it on its own SCSI bus.
The following sections describe how to prepare the TL891 and TL892 in
more detail.
8.6.1 Setting the TL891 or TL892 SCSI ID
The control panel on the front of the TL891 and TL892 MiniLibraries is used
to display power-on self-test (POST) status, display messages, and to set
up MiniLibrary functions.
When power is first applied to a MiniLibrary, a series of POST diagnostics
are performed. During POST execution, the MiniLibrary model number,
current date and time, firmware revision, and the status of each test is
displayed on the control panel.
After the POST diagnostics have completed, the default screen is shown:
DLT0 Idle
DLT1 Idle
Loader Idle
0> _ _ _ _ _ _ _ _ _ _ <9
8–18 Configuring a Shared SCSI Bus for Tape Drive Use
The first and second lines of the default screen show the status of the two
drives (if present). The third line shows the status of the library robotics,
and the fourth line is a map of the magazine, with the numbers from 0 to
9 representing the cartridge slots. Rectangles present on this line indicate
cartridges present in the corresponding slot of the magazine.
For example, this fourth line (0> X X _ _ _ _ _ _ _ <9, where X
represents rectangles) indicates that cartridges are installed in slots 0 and 1.
______________________
Note _______________________
There are no switches for setting a mechanical SCSI ID for the
tape drives. The SCSI IDs default to 5. The MiniLibrary sets the
electronic SCSI ID very quickly, before any device can probe the
MiniLibrary, so the lack of a mechanical SCSI ID does not cause
any problems on the SCSI bus.
To set the SCSI ID, follow these steps:
1. From the Default Screen, press the Enter button to enter the Menu
Mode, displaying the Main Menu.
____________________
Note _____________________
When you enter the Menu Mode, the Ready light goes out, an
indication that the module is off line, and all media changer
commands from the host return a SCSI not ready status
until you exit the Menu Mode and the Ready light comes on
once again.
2. Depress the down arrow button until the Configure Menu item is
selected, then press the Enter button to display the Configure submenu.
____________________
Note _____________________
The control panel up and down arrows have an auto-repeat
feature. When you press either button for more than one-half
second, the control panel behaves as if you were pressing the
button about four times per second. The effect stops when
you release the button.
3. Press the down arrow button until the Set SCSI item is selected and
press the Enter button.
Configuring a Shared SCSI Bus for Tape Drive Use 8–19
4. Select the tape drive (DLT0 Bus ID: or DLT1 Bus ID:) or library robotics
(LIB Bus ID:) for which you wish to change the SCSI bus ID. The default
SCSI IDs are as follows:
•
•
•
Lib Bus ID: 0
DLT0 Bus ID: 4
DLT1 Bus ID: 5
Use the up or down arrow button to select the item for which you need
to change the SCSI ID. Press the Enter button.
5. Use the up or down arrow button to scroll through the possible SCSI ID
settings. Press the Enter button when the desired SCSI ID is displayed.
6. Repeat steps 4 and 5 to set other SCSI bus IDs as necessary.
7. Press the Escape button repeatedly until the default menu is displayed.
8.6.2 Cabling the TL891 or TL892 MiniLibraries
There are six 68-pin, high-density SCSI connectors on the back of the TL891.
The two leftmost connectors are for the library robotics controller. The
middle two are for tape drive 1. The two on the right are for tape drive 2 (if
the TL892 upgrade has been installed).
______________________
Note _______________________
The tape drive SCSI connectors are labeled DLT1 (tape drive 1)
and DLT2 (tape drive 2). The control panel designation for the
drives is DLT0 (tape drive 1) and DLT1 (tape drive 2).
The default for the DLT MiniLibrary TL891 is to place the robotics controller
and tape drive 1 on the same SCSI bus. A 0.3-meter SCSI jumper cable is
provided with the unit. Plug this cable into the second connector (from the
left) and the third connector. If the MiniLibrary has been upgraded to two
drives, place the second drive on the same SCSI bus with another 0.3-meter
SCSI bus jumper cable, or place it on its own SCSI bus.
______________________
Note _______________________
To achieve system performance capabilities, we recommend
placing no more than two TZ89 tape drives on a SCSI bus.
The internal cabling of the TL891 and TL892 is too long to
allow external termination with a trilink/H879-AA combination.
Therefore, the TL891 or TL892 must be the last device on the
shared SCSI bus. They may not be removed from the shared
8–20 Configuring a Shared SCSI Bus for Tape Drive Use
SCSI bus without stopping all ASE services that generate activity
on the bus.
For this reason, we recommend that tape devices be placed on
separate shared SCSI buses, and that there be no storage devices
on the SCSI bus.
The cabling depends on whether or not there are one or two drives, and for
the two-drive configuration, if each drive is on a separate SCSI bus.
______________________
Note _______________________
It is assumed that the library robotics is on the same SCSI bus as
tape drive 1.
To connect the library robotics and one drive to a single shared SCSI bus,
follow these steps:
1. Connect a BN21K or BN21L between the last trilink connector on the
bus to the leftmost connector (as viewed from the rear) of the TL891.
2. Install a 0.3-meter SCSI bus jumper between the rightmost robotics
connector (second connector from the left) and the left DLT1 connector
(the third connector from the left).
3. Install an H879-AA terminator on the right DLT1 connector (the fourth
connector from the left).
To connect the drive robotics and two drives to a single shared SCSI bus,
follow these steps:
1. Connect a BN21K or BN21L between the last trilink connector on the
bus to the leftmost connector (as viewed from the rear) of the TL892.
2. Install a 0.3-meter SCSI bus jumper between the rightmost robotics
connector (the second connector from the left) and the left DLT1
connector (the third connector from the left).
3. Install a 0.3-meter SCSI bus jumper between the rightmost DLT1
connector (the fourth connector from the left) and the left DLT2
connector (the fifth connector from the left).
4. Install an H879-AA terminator on the right DLT2 connector (the
rightmost connector).
Configuring a Shared SCSI Bus for Tape Drive Use 8–21
To connect the drive robotics and one drive to one shared SCSI bus and the
second drive to a second shared SCSI bus, follow these steps:
1. Connect a BN21K or BN21L between the last trilink connector on one
shared SCSI bus to the leftmost connector (as viewed from the rear) of
the TL892.
2. Connect a BN21K or BN21L between the last trilink connector on the
second shared SCSI bus to the left DLT2 connector (the fifth connector
from the left).
3. Install a 0.3-meter SCSI bus jumper between the rightmost robotics
connector (the second connector from the left) and the left DLT1
connector (the third connector from the left).
4. Install an H879-AA terminator on the right DLT1 connector (the fourth
connector from the left) and install another H879-AA terminator on the
right DLT2 connector (the rightmost connector).
Figure 8–9 shows an example of a TruCluster Server cluster with a TL892
connected to two shared SCSI buses.
8–22 Configuring a Shared SCSI Bus for Tape Drive Use
Figure 8–9: TruCluster Server Cluster with a TL892 on Two Shared SCSI
Buses
KZPSA adapter,
trilink connector,
and H879 terminator
DWZZA-VA,
trilink connector,
and H879 terminator
DWZZA-VA,
trilink connector,
and H879 terminator
BN21K or BN21L cables
T
T
T
T
1
2
3
4
5
1
2
3
4
5
Memory
Channel
adapters
Memory Channel link cable
BA350
BA350
AlphaServer 2100A
AlphaServer 2100A
Library
Robotics
DLT1
H879-AA
terminators
BN21K or
BN21L cable
DLT2
Expansion
Unit
Interface
1 Ft
SCSI Bus
Jumper
TL892
ZK-1357U-AI
8.7 Preparing the TL890 DLT MiniLibrary Expansion Unit
The topics in this section provide information on preparing the TL890 DLT
MiniLibrary expansion unit with the TL891 and TL892 DLT MiniLibraries
for use on a shared SCSI bus.
______________________
Note _______________________
To achieve system performance capabilities, we recommend
placing no more than two TZ89 drives on a SCSI bus, and also
recommend that no shared storage be placed on the same SCSI
bus with a tape library.
Configuring a Shared SCSI Bus for Tape Drive Use 8–23
8.7.1 TL890 DLT MiniLibrary Expansion Unit Hardware
The TL890 expansion unit is installed above the TL891/TL892 DLT
MiniLibrary base units in a SW500, SW800, or RETMA cabinet. The
expansion unit integrates the robotics in the individual modules into a
single, coordinated library robotics system. The TL890 assumes control of
the media, maintaining an inventory of all media present in the system, and
controls movement of all media. The tape cartridges can move freely between
the expansion unit and any of the base modules via the system’s robotically
controlled pass-through mechanism. The pass-through mechanism is
attached to the back of the expansion unit and each of the base modules.
For each TL891/TL892 base module beyond the first module, the
pass-through mechanism must be extended by seven inches (the height of
each module) with a DS-TL800-AA pass-through mechanism extension.
A seven-inch gap may be left between base modules (providing there is
sufficient space), but additional pass-through mechanism extensions must
be used.
For complete hardware installation instructions, see the DLT MiniLibrary
(TL890) Expansion Unit User ’s Guide.
The combination of the TL890 expansion unit and the TL891/TL892
MiniLibrary modules is referred to as a DLT MiniLibrary for the remainder
of this discussion.
8.7.2 Preparing the DLT MiniLibraries for Shared SCSI Bus Usage
The following sections describe how to prepare the DLT MiniLibraries in
more detail. It is assumed that the expansion unit, base modules, and
pass-through and motor mechanisms have been installed.
8.7.2.1 Cabling the DLT MiniLibraries
You must make the following connections to render the DLT MiniLibrary
system operational:
•
Expansion unit to the motor mechanism: The motor mechanism cable is
about 1 meter long and has a DB-15 connector on each end. Connect it
between the connector labeled Motor on the expansion unit to the motor
on the pass-through mechanism.
_____________________ Note _____________________
This cable is not shown in Figure 8–10 as the pass-through
mechanism is not shown in the figure.
8–24 Configuring a Shared SCSI Bus for Tape Drive Use
•
Robotics control cables from each base module to the expansion unit:
These cables have a DB-9 male connector on one end and a DB-9 female
connector on the other end. Connect the male end to the Expansion
Unit Interface connector on the base module and the female end to any
Expansion Modules connector on the expansion unit.
_____________________ Note _____________________
It does not matter which interface connector a base module
is connected to.
•
•
SCSI bus connection to the expansion unit robotics: Connect the shared
SCSI bus that will control the robotics to one of the SCSI connectors
on the expansion unit with a BN21K (or BN21L) cable. Terminate the
SCSI bus with an H879-AA terminator on the other expansion unit
SCSI connector.
SCSI bus connection to each of the base module tape drives: Connect a
shared SCSI bus to one of the DLT1 or DLT2 SCSI connectors on each of
the base modules with BN21K (or BN21L) cables. Terminate the other
DLT1 or DLT2 SCSI bus connection with an H879-AA terminator.
You can daisy chain between DLT1 and DLT2 (if present) with a
0.3-meter SCSI bus jumper (supplied with the TL891). Terminate the
SCSI bus at the tape drive on the end of the shared SCSI bus with an
H879-AA terminator.
____________________
Notes
____________________
Do not connect a SCSI bus to the SCSI connectors for the
library connectors on the base modules.
We recommend that no more than two TZ89 tape drives be
on a SCSI bus.
Figure 8–10 shows a MiniLibrary configuration with two TL892 DLT
MiniLibraries and a TL890 DLT MiniLibrary expansion unit. The TL890
library robotics is on one shared SCSI bus, and the two TZ89 tape drives in
each TL892 are on separate, shared SCSI buses. Note that the pass-through
mechanism and cable to the library robotics motor is not shown in this figure.
Configuring a Shared SCSI Bus for Tape Drive Use 8–25
Figure 8–10: TL890 and TL892 DLT MiniLibraries on Shared SCSI Buses
DWZZA-VA,
trilink connector,
and H879 terminator
BN21K or BN21L cables
T
T
T
T
T
1
2
3
4
5
1
2
3
4
5
T
T
Memory
Channel
adapters
Memory Channel link cable
BA350
BA350
AlphaServer 2100As
BN21K or BN21L cables
Diag
Motor
H879-AA
terminator
SCSI
Expansion Modules
TL890
Library
Robotics
DLT2
DLT1
Robotics
Control
cables
H879-AA
terminator
0.3M
SCSI Bus
jumper
TL892
Library
Robotics
DLT2
DLT1
H879-AA
terminator
Expansion
Unit
Interface
TL892
NOTE: This drawing is not to scale.
ZK-1398U-AI
8.7.2.2 Configuring a Base Module as a Slave
The TL891/TL892 base modules are shipped configured as standalone
systems. When they are used in conjunction with the TL890 DLT
MiniLibrary expansion unit, the expansion unit must control the robotics of
each of the base modules. Therefore, the base modules must be configured
as a slave to the expansion unit.
After the hardware and cables are installed, but before you power up
the expansion unit in a MiniLibrary system for the first time, you must
reconfigure each of the base modules in the system as a slave. The expansion
8–26 Configuring a Shared SCSI Bus for Tape Drive Use
unit will not have control over the base module robotics when you power up
the MiniLibrary system if you do not reconfigure the base modules as a slave.
To reconfigure a TL891/TL892 base module as a slave to the TL890 DLT
MiniLibrary expansion unit, perform the following procedure on each base
module in the system:
1. Turn on the power switch on the TL891/TL892 base module to be
reconfigured.
____________________
Note _____________________
Do not power on the expansion unit. Leave it powered off
until all base modules have been reconfigured as slaves.
After a series of power-on self-tests have executed, the default screen
will be displayed on the base module control panel:
DLT0 Idle
DLT1 Idle
Loader Idle
0> _ _ _ _ _ _ _ _ _ _ <9
The default screen shows the state of the tape drives, loader, and
number of cartridges present for this base module. A rectangle in place
of the underscore indicates that a cartridge is present in that location.
2. Press the Enter button to enter the Menu Mode, displaying the Main
Menu.
3. Depress the down arrow button until the Configure Menu item is
selected, then press the Enter button.
____________________
Note _____________________
The control panel up and down arrows have an auto-repeat
feature. When you press either button for more than one-half
second, the control panel behaves as if you were pressing the
button about four times per second. The effect stops when
you release the button.
4. Press the down arrow button until the Set Special Config menu is
selected and press the Enter button.
5. Press the down arrow button repeatedly until the Alternate Config item
is selected and press the Enter button.
6. Press the down arrow button to change the alternate configuration from
the default (Standalone) to Slave. Press the Enter button.
Configuring a Shared SCSI Bus for Tape Drive Use 8–27
7. After the selection stops flashing and the control panel indicates that
the change is not effective until a reboot, press the Enter button.
8. When the Special Configuration menu reappears, turn the power switch
off and then on to cycle the power. The base module is now reconfigured
as a slave to the TL890 expansion unit.
9. Repeat the steps for each TL891/TL892 base module present that is to
be a slave to the TL890 expansion unit.
8.7.2.3 Powering Up the DLT MiniLibrary
When turning on power to the DLT MiniLibrary, power must be applied to
the TL890 expansion unit simultaneously or after power is applied to the the
TL891/TL892 base modules. If the expansion unit is powered on first, its
inventory of modules may be incorrect and the contents of some or all of the
modules will be inacessible to the system and to the host.
When the expansion unit comes up, it will communicate with each base
module through the expansion unit interface and inventory the number of
base modules, tape drives, and cartridges present in each base module. After
the MiniLibrary configuration has been determined, the expansion unit
will communicate with each base module and indicate to the base module
which cartridge group that base module contains. The cartridges slots are
numbered by the expansion unit as follows:
•
•
•
•
Expansion unit: 0 through 15
Top TL891/TL892: 16 through 25
Middle TL891/TL892: 26 through 35
Bottom TL891/TL892: 36 through 45
When all initialization communication between the expansion module
and each base module has completed, the base modules will display their
cartridge numbers according to the remapped cartridge inventory.
For instance, the middle base module default screen would be displayed as
follows:
DLT2 Idle
DLT3 Idle
Loader Idle
26> _ _ _ _ _ _ _ _ _ _ <35
8.7.2.4 Setting the TL890/TL891/TL892 SCSI ID
After the base modules have been reconfigured as slaves, each base module
control panel still provides tape drive status and error information, but all
8–28 Configuring a Shared SCSI Bus for Tape Drive Use
control functions are carried out from the expansion unit control panel. This
includes setting the SCSI ID for each of the tape drives present.
To set the SCSI IDs for the tape drives in a MiniLibrary configured with
TL890/TL891/TL892 hardware, follow these steps:
1. Apply power to the MiniLibrary, ensuring that you power up the
expansion unit after or at the same time as the base modules.
2. Wait until power-on self-tests (POST) have terminated and the
expansion unit and each base module display the default screen.
3. At the expansion unit control panel, press the Enter button to display
the Main Menu.
4. Press the down arrow button until the Configure Menu item is selected,
and then press the Enter button to display the Configure submenu.
5. Press the down arrow button until the Set SCSI item is selected and
press the Enter button.
6. Press the up or down arrow button to select the appropriate tape drive
(DLT0 Bus ID:, DLT1 Bus ID:, DLT2 Bus ID:, and so on) or library
robotics (Library Bus ID:) for which you wish to change the SCSI bus
ID. Assuming that each base module has two tape drives, the top base
module contains DLT0 and DLT1. The next base module down contains
DLT2 and DLT3. The bottom base module contains DLT4 and DLT5.
The default SCSI IDs, after being reconfigured by the expansion unit,
are as follows:
•
•
•
•
•
•
•
Library Bus ID: 0
DLT0 Bus ID: 1
DLT1 Bus ID: 2
DLT2 Bus ID: 3
DLT3 Bus ID: 4
DLT4 Bus ID: 5
DLT5 Bus ID: 6
7. Press Enter when you have the item selected for which you wish to
change the SCSI ID.
8. Use the up and down arrows to select the desired SCSI ID. Press the
Enter button to save the new selection.
9. Press the Escape button once to return to the Set SCSI submenu to
select another tape drive or the library robotics, and then repeat steps 6,
7, and 8 to set the SCSI ID.
Configuring a Shared SCSI Bus for Tape Drive Use 8–29
10. If there are other items you wish to configure, press the Escape button
until the Configure submenu is displayed, then select the item to be
configured. Repeat this procedure for each item you wish to configure.
11. If there are no more items to be configured, press the Escape button
until the Default window is displayed.
8.8 Preparing the TL894 DLT Automated Tape Library for
Shared SCSI Bus Usage
The topics in this section provide information on preparing the TL894 DLT
automated tape library for use on a shared SCSI bus in a TruCluster Server
cluster.
______________________
Note _______________________
To achieve system performance capabilities, we recommend
placing no more than two TZ89 drives on a SCSI bus segment.
We also recommend that storage be placed on shared SCSI buses
that do not have tape drives.
The TL894 midrange automated DLT library contains a robotics controller
and four differential TZ89 tape drives.
The following sections describe how to prepare the TL894 in more detail.
8.8.1 TL894 Robotic Controller Required Firmware
Robotic firmware Version S2.20 is the minimum firmware revision supported
in a TruCluster Server cluster. For information on upgrading the robotic
firmware, see the Flash Download section of the TL81X/ TL894 Automated
Tape Library for DLT Cartridges Diagnostic Software User ’s Manual.
8.8.2 Setting TL894 Robotics Controller and Tape Drive SCSI IDs
The robotics controller, and each tape drive must have the SCSI ID set
(unless the default is sufficient). Table 8–4 lists the default SCSI IDs.
Table 8–4: TL894 Default SCSI ID Settings
SCSI Device
SCSI Address
0
2
3
Robotics Controller
Tape Drive 0
Tape Drive 1
8–30 Configuring a Shared SCSI Bus for Tape Drive Use
Table 8–4: TL894 Default SCSI ID Settings (cont.)
SCSI Device
Tape Drive 2
Tape Drive 3
SCSI Address
4
5
To set the SCSI ID for the TL894 robotics controller, follow these steps:
1. Press and release the Control Panel STANDBY button and verify that
the SDA (Status Display Area) shows System Off-line.
2. Press and release SELECT to enter the menu mode.
3. Verify that the following information is displayed in the SDA:
Menu:
Configuration:
4. Press and release SELECT to choose the Configuration menu.
5. Verify that the following information is displayed in the SDA:
Menu: Configuration
Inquiry
6. Press and release the up or down arrow buttons to locate the SCSI
Address submenu, and verify that the following information is displayed
in the SDA:
Menu: Configuration
SCSI Address ..
7. Press and release the SELECT button to choose the SCSI Address
submenu and verify that the following information is displayed in the
SDA:
Menu: Configuration
Robotics
8. Press and release the SELECT button to choose the Robotics submenu
and verify that the following information is displayed in the SDA:
Menu: SCSI Address
SCSI ID 0
9. Use the up and down arrow buttons to select the desired SCSI ID for the
robotics controller.
10. When the desired SCSI ID is displayed on line 2, press and release
the SELECT button.
11. Press and release the up or down button to clear the resulting display
from the command.
Configuring a Shared SCSI Bus for Tape Drive Use 8–31
12. Press and release the up or down button and the SELECT button
simultaneously, and verify that System On-line or System Off-line is
displayed in the SDA.
To set the SCSI ID for each tape drive if the desired SCSI IDs are different
from those shown in Table 8–4, follow these steps:
1. Press and release the Control Panel STANDBY button and verify that
the SDA (Status Display Area) shows System Off-line.
2. Press and release SELECT to enter the menu mode.
3. Verify that the following information is displayed in the SDA:
Menu:
Configuration:
4. Press and release SELECT to choose the Configuration menu.
5. Verify that the following information is displayed in the SDA:
Menu: Configuration
SCSI Address
6. Press and release the SELECT button again to choose SCSI Address
and verify that the following information is shown in the SDA:
Menu: SCSI Address
Robotics
7. Use the down arrow button to bypass the Robotics submenu and verify
that the following information is shown in the SDA:
Menu: SCSI Address
Drive 0
8. Use the up and down arrow buttons to select the drive number to set
or change.
9. When you have the proper drive number displayed on line 2, press and
release the SELECT button and verify that the following information is
shown in the SDA:
Menu: Drive 0
SCSI ID 0
10. Use the up and down arrow buttons to select the desired SCSI ID for
the selected drive.
11. When the desired SCSI ID is displayed on line 2, press and release
the SELECT button.
12. Repeat steps 8 through 11 to set or change all other tape drive SCSI IDs.
13. Press and release the up or down button to clear the resulting display
from the command.
8–32 Configuring a Shared SCSI Bus for Tape Drive Use
14. Press and release the up or down button and the SELECT button
simultaneously and verify that System On-line or System Off-line
is displayed in the SDA.
8.8.3 TL894 Tape Library Internal Cabling
The default internal cabling configuration for the TL894 tape library has the
robotics controller and top drive (drive 0) on SCSI bus port 1. Drive 1 is on
SCSI bus port 2, drive 2 is on SCSI port 3, and drive 3 is on SCSI bus port 4.
A terminator (part number 0415619) is connected to each of the drives to
provide termination at that end of the SCSI bus.
This configuration, called the four-bus configuration, is shown in
Figure 8–11. In this configuration, each of the tape drives, except SCSI bus
drive 0 and the robotics controller, requires a SCSI address on a separate
SCSI bus. The robotics controller and drive 0 use two SCSI IDs on their
SCSI bus.
Figure 8–11: TL894 Tape Library Four-Bus Configuration
Robotics Controller
*SCSI Address 0
Tape Drive
Interface PWA
SCSI Cable
1.5m
Tape Drive 0
*SCSI Address 2
Internal SCSI
Termination #1
Rear Panel
Host
Tape Drive 1
Connection #4
*SCSI Address 3
Internal SCSI
Termination #2
SCSI Port 4
Rear Panel
Host
Tape Drive 2
*SCSI Address 4
Connection #3
Internal SCSI
Termination #3
Rear Panel
Host
Connection #2
SCSI Port 3
SCSI Port 2
Tape Drive 3
*SCSI Address 5
Rear Panel
Host
Connection #1
Internal SCSI
Termination #4
SCSI Cable
3m
SCSI Port 1
* - Indicates the "default" SCSI ID of the installed devices
ZK-1324U-AI
You can reconfigure the tape drives and robotics controller in a two-bus
configuration by using the SCSI jumper cable (part number 6210567)
supplied in the accessories kit shipped with each TL894 unit. Remove the
terminator from one drive and remove the internal SCSI cable from the
Configuring a Shared SCSI Bus for Tape Drive Use 8–33
other drive to be daisy chained. Use the SCSI jumper cable to connect the
two drives and place them on the same SCSI bus.
______________________ Notes ______________________
We recommend that you not place more than two TZ89 tape
drives on any one SCSI bus in these tape libraries. We also
recommend that storage be placed on shared SCSI buses that
do not have tape drives.
Therefore, we do not recommend that you reconfigure the TL894
tape library into the one-bus configuration.
Appendix B of the TL81X/ TL894 Automated Tape Library
for DLT Cartridges Facilities Planning and Installation Guide
provides figures showing various bus configurations. In these
figures, the configuration changes have been made by removing
the terminators from both drives, installing the SCSI bus jumper
cable on the drive connectors vacated by the terminators, then
installing an HD68 SCSI bus terminator on the SCSI bus port
connector on the cabinet exterior.
This is not wrong, but by reconfiguring in this manner, the
length of the SCSI bus is increased by 1.5 meters, and may cause
problems if SCSI bus length is of concern.
In a future revision of the previously mentioned guide, the bus
configuration figures will be modified to show all SCSI buses
terminated at the tape drives.
8.8.4 Connecting the TL894 Tape Library to the Shared SCSI Bus
The TL894 tape libraries have up to 3 meters of internal SCSI cabling per
SCSI bus. Because of the internal SCSI cable lengths, it is not possible to
use a trilink connector or Y cable to terminate the SCSI bus external to the
library as is done with other devices on the shared SCSI bus. Each SCSI bus
must be terminated internal to the tape library, at the tape drive itself with
the installed SCSI terminators. Therefore, TruCluster Server clusters using
the TL894 tape library must ensure that the tape library is on the end of
the shared SCSI bus.
In a TruCluster Server cluster with a TL894 tape library, the member
systems and StorageWorks enclosures or RAID subsystems may be isolated
from the shared SCSI bus because they use trilink connectors or Y cables.
However, the ASE must be shut down to remove a tape loader from the
shared bus.
8–34 Configuring a Shared SCSI Bus for Tape Drive Use
Figure 8–12 shows a sample TruCluster Server cluster using a TL894 tape
library. In the sample configuration, the tape library has been connected in
the two-bus mode by jumpering tape drive 0 to tape drive 1 and tape drive
2 to tape drive 3 (See Section 8.8.3 and Figure 8–11). The two SCSI buses
are left at the default SCSI IDs and terminated at drives 1 and 3 with the
installed terminators (part number 0415619).
To add a TL894 to a shared SCSI bus, select the member system or storage
device that will be the next to last device on the shared SCSI bus. Connect a
BN21K or BN21L cable between the Y cable on that device to the appropriate
tape library port. In Figure 8–12, one bus is connected to port 1 (robotics
controller and tape drives 0 and 1) and the other bus is connected to port
3 (tape drives 2 and 3). Ensure that the terminators are present on the
tape drives 1 and 3.
Figure 8–12: Shared SCSI Buses with TL894 in Two-Bus Mode
Network
Memory
Channel
Interface
Member System 1
Member System 2
7
7
Memory Channel
Memory Channel
T
6
5
KZPBA-CB (ID 6)
KZPBA-CB (ID 6)
KZPBA-CB (ID 6)
KZPBA-CB (ID 7)
5
T
KZPBA-CB (ID 7)
KZPBA-CB (ID 7)
5
5
T
T
7
7
1
1
T
DS-DWZZH-03
T
T
(2-bus mode)
2
3
4
T
SCSI Port 4
Controller B
HSZ70
Controller A
HSZ70
SCSI Port 3
SCSI Port 2
SCSI Port 1
StorageWorks
RAID Array 7000
TL894
NOTE: This drawing is not to scale.
ZK-1625U-AI
Configuring a Shared SCSI Bus for Tape Drive Use 8–35
8.9 Preparing the TL895 DLT Automated Tape Library for
Shared SCSI Bus Usage
The topics in this section provide information on preparing the TL895 Digital
Linear Tape (DLT) automated tape library for use on a shared SCSI bus.
______________________
Note _______________________
To achieve system performance capabilities, we recommend
placing no more than two TZ89 drives on a SCSI bus segment. We
also recommend that storage be placed on shared SCSI buses that
do not have tape drives. This makes it easier to stop ASE services
affecting the SCSI bus that the tape loaders are on.
The DS-TL895-BA automated digital linear tape library consists of five
TZ89N-AV tape drives and 100 tape cartridge bins (96 storage bins in a
fixed-storage array (FSA) and 4 load port bins). The storage bins hold either
CompacTape III, CompacTape IIIXT, or CompacTape IV cartridges. The
maximum storage capacity of the library is 3500 GB uncompressed, based
upon 100 CompacTape IV cartridges at 35 GB each. For more information on
the TL895, see the following manuals:
•
TL895 DLT Tape Library Facilities Planning and Installation Guide
(EK-TL895-IG)
•
•
TL895 DLT Library Operator ’s Guide (EK-TL895-OG)
TL895 DLT Tape Library Diagnostic Software User ’s Manual
(EK-TL895-UM)
For more information on upgrading from five to six or seven tape drives, see
the TL895 Drive Upgrade Instructions manual.
______________________
Note _______________________
There are rotary switches on the library printed circuit board
used to set the library and tape drive SCSI IDs. The SCSI IDs
set by these switches are used for the first 20 to 30 seconds after
power is applied, until the electronics is activated and able to
set the SCSI IDs electronically.
The physical SCSI IDs should match the SCSI IDs set by the
library electronics. Ensure that the SCSI ID set by the rotary
switch and from the control panel do not conflict with any SCSI
bus controller SCSI ID.
The following sections describe how to prepare the TL895 for use on a shared
SCSI bus in more detail.
8–36 Configuring a Shared SCSI Bus for Tape Drive Use
8.9.1 TL895 Robotic Controller Required Firmware
Robotic firmware version N2.20 is the minimum firmware revision supported
in a TruCluster Server cluster. For information on upgrading the robotic
firmware, see the Flash Download section of the TL895 DLT Tape Library
Diagnostic Software User ’s Manual.
8.9.2 Setting the TL895 Tape Library SCSI IDs
The library and each tape drive must have the SCSI ID set (unless the
default is sufficient). Table 8–5 lists the TL895 default SCSI IDs.
Table 8–5: TL895 Default SCSI ID Settings
SCSI Device
Library
Drive 0
Drive 1
Drive 2
Drive 3
Drive 4
Drive 5
Drive 6
SCSI ID
0
1
2
3
4
5
1
2
The SCSI IDs must be set mechanically by the rotary switches, and
electronically from the control panel. After you have set the SCSI IDs from
the switches, power up the library and electronically set the SCSI IDs.
To electronically set the SCSI ID for the TL895 library and tape drives,
follow these steps:
1. At the control panel, press the Operator tab.
2. On the Enter Password screen, enter the operator password. The
default operator password is 1234. The lock icon is unlocked and shows
an O to indicate that you have operator-level security clearance.
3. On the Operator screen, press the Configure Library button. The
Configure Library screen displays the current library configuration.
____________________
Note _____________________
You can configure the library model number, number of
storage bins, number of drives, library SCSI ID, and tape
drive SCSI IDs from the Configure Library screen.
Configuring a Shared SCSI Bus for Tape Drive Use 8–37
4. To change any of the configurations, press the Configure button.
5. Press the Select button until the item you wish to configure is
highlighted. For the devices, select the desired device (library or drive)
by scrolling through the devices with the arrow buttons. After the
library or selected drive is selected, use the Select button to highlight
the SCSI ID.
6. Use the arrow buttons to scroll through the setting choices until the
desired setting appears.
7. When you have the desired setting, press the Change button to save the
setting as part of the library configuration.
8. Repeat steps 5 through 7 to make additional changes to the library
configuration.
9. Place the library back at the user level of security as follows:
•
•
•
Press the lock icon on the vertical bar of the control panel.
On the Password screen, press the User button.
A screen appears informing you that the new security level has
been set. Press the Okay button. The lock icon appears as a locked
lock and displays a U to indicate that the control panel is back at
User level.
10. Power cycle the tape library to allow the new SCSI IDs to take affect.
8.9.3 TL895 Tape Library Internal Cabling
The default internal cabling configuration for the TL895 tape library has
the library robotics controller and top drive (drive 0) on SCSI bus port 1.
Drive 1 is on SCSI bus port 2, drive 2 is on SCSI bus port 3, and so on. A
terminator (part number 0415619) is connected to each of the drives to
provide termination at the tape drive end of the SCSI bus.
In this configuration each of the tape drives, except tape drive 0 and the
robotics controller, require a SCSI ID on a separate SCSI bus. The robotics
controller and tape drive drive 0 use two SCSI IDs on their SCSI bus.
You can reconfigure the tape drives and robotics controller to place multiple
tape drives on the same SCSI bus with SCSI bus jumper (part number
6210567) included with the tape library.
______________________
Note _______________________
We recommend placing no more than two TZ89 drives on a SCSI
bus segment. We also recommend that storage be placed on
shared SCSI buses that do not have tape drives.
8–38 Configuring a Shared SCSI Bus for Tape Drive Use
To reconfigure TL895 SCSI bus configuration, follow these steps:
1. Remove the SCSI bus cable from one drive to be daisy chained.
2. Remove the terminator from the other drive to be daisy chained.
3. Ensure that the drive that will be the last drive on the SCSI bus has a
terminator installed.
4. Install a SCSI bus jumper cable (part number 6210567) on the open
connectors of the two drives to be daisy chained.
Figure 8–13 shows an example of a TL895 that has tape drives 1, 3, and 5
daisy chained to tape drives 2, 4, and 6 respectively.
Figure 8–13: TL895 Tape Library Internal Cabling
Robotics
Controller
SCSI ID 0
Tape Drive 0
Terminator
SCSI ID 1
PN 0415619
Tape Drive 1
SCSI ID 2
SCSI Jumper Cable
PN 6210567
SCSI Port 8
SCSI Port 7
Tape Drive 2
SCSI ID 3
Terminator
SCSI Port 6
SCSI Port 5
SCSI Port 4
SCSI Port 3
Tape Drive 3
SCSI ID 4
Jumper
Cable
Tape Drive 4
SCSI ID 5
Terminator
Tape Drive 5
SCSI ID 1
SCSI Port 2
SCSI Port 1
Jumper
Cable
Tape Drive 6
SCSI ID 2
Terminator
ZK-1397U-AI
Configuring a Shared SCSI Bus for Tape Drive Use 8–39
8.9.4 Upgrading a TL895
The TL985 DLT automated tape library can be upgraded from two or
five tape drives to seven drives with multiple DS-TL89X-UA upgrade
kits. Besides the associated documentation, the upgrade kit contains one
TZ89N-AV tape drive, a SCSI bus terminator, a SCSI bus jumper (part
number 6210567) so you can place more than one drive on the same SCSI
bus, and other associated hardware.
Before the drive is physically installed, set the SCSI ID rotary switches
(on the library printed circuit board) to the same SCSI ID that will be
electronically set. After the drive installation is complete, set the electronic
SCSI ID using the Configure menu from the control panel (see Section 8.9.2).
The actual upgrade is beyond the scope of this manual. See the TL895 Drive
Upgrade Instructions manual for upgrade instructions.
8.9.5 Connecting the TL895 Tape Library to the Shared SCSI Bus
The TL895 tape library has up to 3 meters of internal SCSI cabling per SCSI
bus. Because of the internal SCSI cable lengths, it is not possible to use a
trilink connector or Y cable to terminate the SCSI bus external to the library
as is done with other devices on the shared SCSI bus. Each SCSI bus must
be terminated internal to the tape library at the tape drive itself with the
installed SCSI terminators. Therefore, TruCluster Server clusters using the
TL895 tape libraries must ensure that the tape libraries are on the end of
the shared SCSI bus.
In a TruCluster Server cluster with a TL895 tape library, the member
systems and StorageWorks enclosures or RAID subsystems may be isolated
from the shared SCSI bus because they use trilink connectors or Y cables.
However, because the TL895 cannot be removed from the shared SCSI bus,
all ASE services that use any shared SCSI bus attached to the TL895 must
be stopped before the tape loader can be removed from the shared bus.
To add a TL895 tape library to a shared SCSI bus, select the member system
or storage device that will be the next to last device on the shared SCSI bus.
Connect a BN21K or BN21L cable between a trilink or Y cable on that device
to the appropriate tape library port.
8.10 Preparing the TL893 and TL896 Automated Tape
Libraries for Shared SCSI Bus Usage
The topics in this section provide information on preparing the TL893 and
TL896 Automated Tape Libraries (ATLs) for use on a shared SCSI in a
TruCluster Server cluster.
8–40 Configuring a Shared SCSI Bus for Tape Drive Use
______________________
Note _______________________
To achieve system performance capabilities, We recommend
placing no more than two TZ89 drives on a SCSI bus.
The TL893 and TL896 Automated Tape Libraries (ATLs) are designed to
provide high-capacity storage and robotic access for the Digital Linear Tape
(DLT) series of tape drives. They are identical except in the number of tape
drives and the maximum capacity for tape cartridges.
Each tape library comes configured with a robotic controller and bar code
reader (to obtain quick and accurate tape inventories).
The libraries have either three or six TZ89N-AV drives. The TL896, because
it has a greater number of drives, has a lower capacity for tape cartridge
storage.
Each tape library utilizes bulk loading of bin packs, with each bin pack
containing a maximum of 11 cartridges. Bin packs are arranged on an
eight-sided carousel that provides either two or three bin packs per face. A
library with three drives has a carousel three bin packs high. A library with
six drives has a carousel that is only two bin packs high. This provides for
a total capacity of 24 bin packs (264 cartridges) for the TL893, and 16 bin
packs (176 cartridges) for the TL896.
The tape library specifications are as follows:
•
TL893 — The TL893 ATL is a high-capacity, 264-cartridge tape library
providing up to 18.4 TB of storage. The TL893 uses three fast-wide,
differential TZ89N-AV DLT tape drives. It has a maximum transfer rate
of almost 10 MB per second (compressed) for each drive, or a total of
about 30 MB per second.
The TL893 comes configured for three SCSI-2 buses (a three-bus
configuration). The SCSI bus connector is high-density 68-pin,
differential.
•
TL896 — The TL896 ATL is a high-capacity, 176-cartridge tape library
providing up to 12.3 TB of storage. The TL896 uses six fast-wide,
differential TZ89N-AV DLT tape drives. It also has a maximum transfer
rate of almost 10 MB per second per drive (compressed), or a total of
about 60 MB per second.
The TL896 comes configured for six SCSI-2 buses (a six-bus
configuration). The SCSI bus connector is also high-density 68-pin,
differential.
Both the TL893 and TL896 can be extended by adding additional cabinets
(DS-TL893-AC for the TL893 or DS-TL896-AC for the TL896). See the
TL82X Cabinet-to-Cabinet Mounting Instructions manual for information
Configuring a Shared SCSI Bus for Tape Drive Use 8–41
on adding additional cabinets. Up to five cabinets are supported with the
TruCluster Server.
For TruCluster Server, the tape cartridges in all the cabinets are combined
into one logical unit, with consecutive numbering from the first cabinet to
the last cabinet, by an upgrade from the multi-unit, multi-LUN (MUML)
configuration to a multi-unit, single-LUN (MUSL) configuration. See
the TL82X/ TL89X MUML to MUSL Upgrade Instructions manual for
information on the firmware upgrade.
These tape libraries each have a multi-unit controller (MUC) that serves
two functions:
•
It is a SCSI adapter that allows the SCSI interface to control
communications between the host and the tape library.
•
It permits the host to control up to five attached library units in a
multi-unit configuration. Multi-unit configurations are not discussed in
this manual. For more information on multi-unit configurations, see
the TL82X/ TL893/ TL896 Automated Tape Library for DLT Cartridges
Facilities Planning and Installation Guide.
The following sections describe how to prepare these tape libraries in more
detail.
8.10.1 Communications with the Host Computer
Two types of communications are possible between the tape library and
the host computer: SCSI and EIA/TIA-574 serial (RS-232 for nine-pin
connectors). Either method, when used with the multi-unit controller
(MUC), allows a single host computer to control up to five units.
A TruCluster Server cluster supports SCSI communications only between
the host computer and the MUC. With SCSI communications, both control
signals and data flow between the host computer and tape library use the
same SCSI cable. The SCSI cable is part of the shared SCSI bus.
An RS-232 loopback cable must be connected between the Unit 0 and Input
nine-pin connectors on the rear connector panel. The loopback cable connects
the MUC to the robotic controller electronics.
Switch 7 on the MUC switch pack must be down to select the SCSI bus.
8.10.2 MUC Switch Functions
Switch pack 1 on the rear of the multi-unit controller (MUC) is located
below the MUC SCSI connectors. The switches provide the functions shown
in Table 8–6.
8–42 Configuring a Shared SCSI Bus for Tape Drive Use
Table 8–6: MUC Switch Functions
Function
Switch
1, 2, and 3
MUC SCSI ID if Switch 7 is downa
Must be down, reserved for testing
Default is up, disable bus reset on power up
Host selection: Down for SCSI, up for seriala
Must be down, reserved for testing
4 and 5
6
7
8
a
For a TruCluster Server cluster, switch 7 is down, allowing switches 1, 2, and 3 to select the MUC SCSI ID.
8.10.3 Setting the MUC SCSI ID
The multi-unit controller (MUC) SCSI ID is set with switch 1, 2, and 3, as
shown in Table 8–7. Note that switch 7 must be down to select the SCSI bus
and enable switches 1, 2, and 3 to select the MUC SCSI ID.
Table 8–7: MUC SCSI ID Selection
MUC SCSI ID
SW1
Down
Up
SW2
Down
Down
Up
SW3
Down
Down
Downa
Down
Up
0
1
2
3
4
5
6
7
Down
Up
Up
Down
Up
Down
Down
Up
Up
Down
Up
Up
Up
Up
a
This is the default MUC SCSI ID.
8.10.4 Tape Drive SCSI IDs
Each tape library arrives with default SCSI ID selections. The TL893 is
shown in Table 8–8. The TL896 is shown in Table 8–9.
If you must modify the tape drive SCSI IDs, use the push-button up-down
counters on the rear of the drive to change the SCSI ID.
Configuring a Shared SCSI Bus for Tape Drive Use 8–43
Table 8–8: TL893 Default SCSI IDs
Device
SCSI Port
Default SCSI ID
2
5
4
3
MUC
C
Drive 2 (top)
Drive 1 (middle)
Drive 0 (bottom)
B
A
Table 8–9: TL896 Default SCSI IDs
Device
SCSI Port
Default SCSI ID
2
5
4
3
5
4
3
MUC
D
Drive 5 (top)
Drive 4
E
F
Drive 3
A
B
C
Drive 2
Drive 1
Drive 0 (bottom)
8.10.5 TL893 and TL896 Automated Tape Library Internal Cabling
The default internal cabling configurations for the TL893 and TL896
Automated Tape Libraries (ATLs) is as follows:
•
The SCSI input for the TL893 is high-density, 68-pin differential. The
default internal cabling configuration for the TL893 is a three-bus mode
shown in Figure 8–14 as follows:
–
The top shelf tape drive (SCSI ID 5) and MUC (SCSI ID 2) are on
SCSI Port C and are terminated on the MUC. To allow the use of
the same MUC and terminator used with the TL822 and TL826, a
68-pin to 50-pin adapter is used on the MUC to connect the SCSI
cable from the tape drive to the MUC. In Figure 8–14 it is shown as
part number 0425031, the SCSI Diff Feed Through. This SCSI bus
is terminated on the MUC with terminator part number 0415498, a
50-pin Micro-D terminator.
–
The middle shelf tape drive (SCSI ID 4) is on SCSI Port B and is
terminated on the drive with a 68-pin Micro-D terminator, part
number 0415619.
8–44 Configuring a Shared SCSI Bus for Tape Drive Use
–
The bottom shelf tape drive (SCSI ID 3) is on SCSI Port A and is
also terminated on the drive with a 68-pin Micro-D terminator, part
number 0415619.
Figure 8–14: TL893 Three-Bus Configuration
0415498 (50-Pin Micro-D Terminator)
0425031 (SCSI Diff Feed Through)
MUC
0425017 (Cable)
SCSI Address 2
TZ89 Tape Drive
SCSI Address 5
(top shelf)
0415619
(68-pin Micro-D Terminator)
TZ89 Tape Drive
SCSI Address 4
(middle shelf)
0415619
(68-pin Micro-D Terminator)
TZ89 Tape Drive
SCSI Address 3
(bottom shelf)
Drive Housing
SCSI Port A
SCSI Port B
SCSI Port C
(Rear Connector Panel)
ZK-1326U-AI
•
The SCSI input for the TL896 is also high-density, 68-pin differential.
The default internal cabling configuration for the TL896 is a six-bus
configuration shown in Figure 8–15 as follows:
–
The upper bay top shelf tape drive (tape drive 5, SCSI ID 5) and
MUC (SCSI ID 2) are on SCSI Port D. To allow the use of the same
MUC and terminator used with the TL822 and TL826, a 68-pin to
50-pin adapter is used on the MUC to connect the SCSI cable from
the tape drive to the MUC. In Figure 8–15 it is shown as part number
0425031, SCSI Diff Feed Through. This SCSI bus is terminated on
the MUC with terminator part number 0415498, a 50-pin Micro-D
terminator.
–
–
The upper bay middle shelf tape drive (tape drive 4, SCSI ID 4) is on
SCSI Port E and is terminated on the tape drive.
The upper bay bottom shelf tape drive (tape drive 3, SCSI ID 3) is on
SCSI Port F and is terminated on the tape drive.
Configuring a Shared SCSI Bus for Tape Drive Use 8–45
–
–
–
–
The lower bay top shelf tape drive (tape drive 2, SCSI ID 5) is on
SCSI Port A and is terminated on the tape drive.
The lower bay middle shelf tape drive (tape drive 1, SCSI ID 4) is on
SCSI Port B and is terminated on the tape drive.
The lower bay bottom shelf tape drive (tape drive 0, SCSI ID 3) is on
SCSI Port C and is terminated on the tape drive.
The tape drive terminators are 68-pin differential terminators (part
number 0415619).
Figure 8–15: TL896 Six-Bus Configuration
0415498 (50-Pin Micro-D Terminator)
0425031 (SCSI Diff Feed Through)
0425017 (Cable)
MUC
SCSI Address 2
TZ89 Drive 5
SCSI Address 5
(top shelf)
TZ89 Drive 4
SCSI Address 4
(middle shelf)
Upper
Bay
0415619
(68-pin Terminator)
TZ89 Drive 3
SCSI Address 3
(bottom shelf)
0415619
(68-pin Terminator)
TZ89 Drive 2
SCSI Address 5
(top shelf)
0415619
(68-pin Terminator)
TZ89 Drive 1
SCSI Address 4
(middle shelf)
Lower
Bay
0415619
(68-pin Terminator)
TZ89 Drive 0
SCSI Address 3
(bottom shelf)
0415619
(68-pin Terminator)
SCSI Port A
SCSI Port D
SCSI Port G
SCSI Port H
(Rear Connector Panel)
SCSI Port B
SCSI Port C
SCSI Port E
SCSI Port F
SCSI Port I
ZK-1327U-AI
8–46 Configuring a Shared SCSI Bus for Tape Drive Use
8.10.6 Connecting the TL893 and TL896 Automated Tape Libraries to
the Shared SCSI Bus
The TL893 and TL896 Automated Tape Libraries (ATLs) have up to 3
meters of internal SCSI cabling on each SCSI bus. Because of the internal
SCSI cable lengths, it is not possible to use a trilink connector or Y cable to
terminate the SCSI bus external to the library as is done with other devices
on the shared SCSI bus. Each SCSI bus must be terminated internal to the
tape library at the tape drive itself with the installed SCSI terminators.
Therefore, TL893 and TL896 tape libraries must be on the end of the shared
SCSI bus.
In a TruCluster Server cluster with TL893 or TL896 tape libraries, the
member systems and StorageWorks enclosures or RAID subsystems may
be isolated from the shared SCSI bus because they use trilink connectors
or Y cables. However, if there is disk storage and an ATL on the same
shared SCSI bus, the ASE must be shut down to remove a tape library from
the shared bus.
You can reconfigure the tape drives and robotics controller to generate other
bus configurations by using the jumper cable (ATL part number 0425017)
supplied in the accessories kit shipped with each TL893 or TL896 unit.
Remove the terminator from one drive and remove the internal SCSI cable
from the other drive to be daisy chained. Use the jumper cable to connect the
two drives and place them on the same SCSI bus.
______________________
Note _______________________
We recommend that you not place more than two drives on any
one SCSI bus in these tape libraries.
Figure 8–16 shows a sample TruCluster Server cluster using a TL896 tape
library in a three-bus configuration. In this configuration, tape drive 4 (Port
E) has been jumpered to tape drive 5, tape drive 2 (Port A) has been jumpered
to tape drive 3, and tape drive 1 (Port B) has been jumpered to tape drive 0.
To add a TL893 or TL896 tape library to a shared SCSI bus, select the
member system that will be the next to the last device on the shared SCSI
bus (the tape library always has to be the last device on the shared SCSI
bus). Connect a BN21K, BN21L, or BN31G cable between the Y cable on
the SCSI bus controller on that member system and the appropriate tape
library port. In Figure 8–16, one shared SCSI bus is connected to port
B (tape drives 0 and 1), one shared SCSI bus is connected to port A (tape
drives 2 and 3), and a third shared SCSI bus is connected to port E (tape
drives 4 and 5 and the MUC).
Configuring a Shared SCSI Bus for Tape Drive Use 8–47
Figure 8–16: Shared SCSI Buses with TL896 in Three-Bus Mode
Network
7
Memory
Channel
Interface
Member System 1
Member System 2
7
Memory Channel
KZPBA-CB (ID 6)
KZPBA-CB (ID 6)
KZPBA-CB (ID 6)
Memory Channel
KZPBA-CB (ID 7)
KZPBA-CB (ID 7)
KZPBA-CB (ID 7)
6
T
6
5
T
6
5
5
5
T
5
5
T
KZPBA-CB (ID 6)
T
KZPBA-CB (ID 7)
7
1
1
T
DS-DWZZH-03
T
T
7
2
3
7
4
T
Controller B
HSZ70
Controller A
HSZ70
StorageWorks
D
E
F
A
B
C
TL896
RAID Array 7000
SCSI Ports
(3-bus mode)
NOTE: This drawing is not to scale.
ZK-1626U-AI
8.11 Preparing the TL881 and TL891 DLT MiniLibraries for
Shared Bus Usage
The topics in this section provide an overview of the Compaq StorageWorks
TL881 and TL891 Digital Linear Tape (DLT) MiniLibraries and hardware
configuration information for preparing the TL881 or TL891 DLT
MiniLibrary for use on a shared SCSI bus.
8.11.1 TL881 and TL891 DLT MiniLibraries Overview
For more information on the TL881 or TL891 DLT MiniLibraries, see the
following Compaq documentation:
•
TL881 MiniLibrary System User ’s Guide
8–48 Configuring a Shared SCSI Bus for Tape Drive Use
•
•
•
TL891 MiniLibrary System User ’s Guide
TL881 MiniLibrary Drive Upgrade Procedure
Pass-Through Expansion Kit Installation Instructions
The TL881 and TL891 Digital Linear Tape (DLT) MiniLibraries are offered
as standalone tabletop units or as expandable rackmount units.
The following sections describe these units in more detail.
8.11.1.1 TL881 and TL891 DLT MiniLibrary Tabletop Model
The TL881 and TL891 DLT MiniLibrary tabletop model consists of one unit
with a removable 10-cartridge magazine, integral bar code reader, and either
one or two DLT 20/40 (TL881) or DLT 35/70 (TL891) drives.
The TL881 DLT MiniLibrary tabletop model is available as either fast,
wide differential or fast, wide single-ended. The single-ended model is not
supported in a TruCluster Server configuration.
The TL891 DLT MiniLibrary tabletop model is only available as fast, wide
differential.
8.11.1.2 TL881 and TL891 MiniLibrary Rackmount Components
A TL881 or TL891 base unit (which contains the tape drive(s)) can operate
as an independent, standalone unit, or in concert with an expansion unit
and multiple data units.
A rackmount multiple-module configuration is expandable to up to six
modules in a configuration. The configuration must contain at least one
expansion unit and one base unit. The TL881 and TL891 DLT MiniLibraries
may include various combinations of:
•
MiniLibrary Expansion unit — the MiniLibrary expansion unit enables
multiple TL881 or TL891 modules to share data cartridges and work as a
single virtual library. The expansion unit also includes a 16-cartridge
magazine.
The expansion unit integrates the robotics in the individual modules into
a single coordinated library robotics system. The expansion unit assumes
control of the media, maintaining an inventory of all media present in
the system, and controls movement of all media. The tape cartridges can
move freely between the expansion unit and any of the base units or data
units via the system’s robotically controlled pass-through mechanism.
The expansion unit can control up to five additional attached modules
(base units and data units) to create a multimodule rackmount
configuration. The expansion unit must be enabled to control the base
unit by setting the base unit to slave mode. The data unit is a passive
Configuring a Shared SCSI Bus for Tape Drive Use 8–49
device and only works as a slave to the expansion unit. To create a
multimodule rackmount system, there must be one expansion unit and
at least one base unit. The expansion unit has to be the top module
in the configuration.
The expansion unit works with either the TL881 or TL891 base unit.
•
TL881 or TL891 base unit — includes library robotics, bar code reader, a
removable 10-cartridge magazine, and one or two tape drives:
–
–
TL881 — DLT 20/40 (TZ88N-AV) drives
TL891 — DLT 35/70 (TZ89N-AV) drives
To participate in a MiniLibrary configuration, each base unit must be
set up as a slave unit to pass control to the expansion unit. Once the
expansion unit has control over the base unit, the expansion unit controls
tape-cartridge movement between the magazines and tape drives.
_____________________ Note _____________________
You cannot mix TL881 and TL891 base units in a rackmount
configuration as the tape drives use different formats.
•
•
Data unit — This rackmount module contains a 16-cartridge magazine
to provide additional capacity in a multi-module configuration. The data
unit robotics works in conjunction with the robotics of the expansion unit
and base units. It is under control of the expansion unit.
The data unit works with either the TL881 or TL891 base unit.
Pass through mechanism — The pass-through mechanism is attached
to the back of the expansion unit and each of the other modules and
allows the transfer of tape cartridges between the various modules. It is
controlled by the expansion unit.
For each base or data unit added to a configuration, the pass-through
mechanism must be extended by seven inches (the height of each
module). A seven-inch gap may be left between modules (providing there
is sufficient space), but additional pass-through mechanism extensions
must be used.
8.11.1.3 TL881 and TL891 Rackmount Scalability
The rackmount version of the TL881 and TL891 MiniLibraries provides
a scalable tape library system that you can configure for maximum
performance, maximum capacity, or various combinations between the
extremes.
Either library uses DLT IV tape cartridges but can also use DLT III or DLT
IIIxt tape cartridges. Table 8–10 shows the capacity and performance of a
8–50 Configuring a Shared SCSI Bus for Tape Drive Use
TL881 or TL891 MiniLibrary in configurations set up for either maximum
performance or maximum capacity.
Table 8–10: TL881 and TL891 MiniLibrary Performance and Capacity Comparison
TL881 MiniLibrary
TL891 MiniLibrary
Number of
Number of Transfer
Transfer
Ratef
Configured
for
Storage
Storage
Base Unitsa b Data Unitsc Rated
Capacitye
Capacityg
Maximum:
5
1
0
4
Performance
15 MB/sec 1.32 TB (66 50 MB/sec 2.31 TB (66
(54
GB/hour)
cartridges) (180
GB/hour)
cartridges)
Capacity
3 MB/sec
(10.8
GB/hour)
1.8 TB (90 10 MB/sec 3.15 TB (90
cartridges) (36 cartridges)
GB/hour)
a
Using an expansion unit with a full 16-cartridge magazine.
Each base unit has a full 10-cartridge magazine and two tape drives.
Using a data unit with full 16-cartridge magazine.
Up to 1.5 MB/sec per drive.
Based on 20 GB/cartridge uncompressed. It could be up to 40 GB/cartridge compressed.
Up to 5 MB/sec per drive.
b
c
d
e
f
g
Based on 35 GB/cartridge uncompressed. It could be up to 70 GB/cartridge compressed.
By modifying the combinations of base units and data units, the performance
and total capacity can be adjusted to meet the customers’ needs.
8.11.1.4 DLT MiniLibrary Part Numbers
Table 8–11 shows the part numbers for the TL881 and TL891 DLT
MiniLibrary systems. Part numbers are only shown for the TL881 fast,
wide differential components.
Table 8–11: DLT MiniLibrary Part Numbers
Number of Tape
Drives
Part Number
DLT Library Component
Tabletop/Rackmount
1
2
1
TL881 DLT Library
TL881 DLT Library
Tabletop
128667-B21
128667-B22
128669-B21
Tabletop
TL881 DLT MiniLibrary
Base Unit
Rackmount
2
1
TL881 DLT MiniLibrary
Base Unit
Rackmount
N/A
128669-B22
128671-B21
Add-on DLT 20/40 drive
for TL881
1
2
TL891 DLT Library
TL891 DLT Library
Tabletop
Tabletop
120875-B21
120875-B22
Configuring a Shared SCSI Bus for Tape Drive Use 8–51
Table 8–11: DLT MiniLibrary Part Numbers (cont.)
Number of Tape
Drives
Part Number
120876-B21
120876-B22
120878-B21
DLT Library Component
Tabletop/Rackmount
Rackmount
Rackmount
N/A
1
2
1
TL891 DLT MiniLibrary
Base Unit
TL891 DLT MiniLibrary
Base Unit
Add-on DLT 35/70 drive
for TL891
MiniLibrary Expansion Unit
MiniLibrary Data Unit
N/A
N/A
Rackmount
Rackmount
120877-B21
128670-B21
______________________
Note _______________________
The TL881 DLT MiniLibrary tabletop model is available as fast,
wide differential or fast, wide single-ended. The single-ended
model is not supported in a cluster configuration. The TL891
DLT MiniLibrary tabletop model is only available as fast, wide
differential.
8.11.2 Preparing a TL881 or TL891 MiniLibrary for Shared SCSI Bus
Use
The following sections describe how to prepare the TL881 and TL891 DLT
MiniLibraries for shared SCSI bus use in more detail.
8.11.2.1 Preparing a Tabletop Model or Base Unit for Standalone Shared SCSI
Bus Usage
A TL881 or TL891 DLT MiniLibrary tabletop model or a rackmount base
unit may be used standalone. You may want to purchase a rackmount base
unit for future expansion.
______________________
Note _______________________
To achieve system performance capabilities, we recommend
placing no more than two tape drives on a SCSI bus, and also
recommend that no shared storage be placed on the same SCSI
bus with a tape library.
8–52 Configuring a Shared SCSI Bus for Tape Drive Use
The topics in this section provide information on preparing the TL881 or
TL891 DLT MiniLibrary tabletop model or rackmount base unit for use
on a shared SCSI bus.
For complete hardware installation instructions, see the TL881 MiniLibrary
System User ’s Guide or TL891 MiniLibrary System User ’s Guide.
8.11.2.1.1 Setting the Standalone MiniLibrary Tape Drive SCSI ID
The control panel on the front of the TL891 and TL892 MiniLibraries is used
to display power-on self-test (POST) status, display messages, and to set
up MiniLibrary functions.
When power is first applied to a MiniLibrary, a series of POST diagnostics
are performed. During POST execution, the MiniLibrary model number,
current date and time, firmware revision, and the status of each test is
displayed on the control panel.
After the POST diagnostics have completed, the default screen is shown:
DLT0 Idle
DLT1 Idle
Loader Idle
0> _ _ _ _ _ _ _ _ _ _ <9
The first and second lines of the default screen show the status of the two
(if present) drives. The third line shows the status of the library robotics,
and the fourth line is a map of the magazine, with the numbers from 0 to
9 representing the cartridge slots. Rectangles present on this line indicate
cartridges present in the corresponding slot of the magazine.
For example, this fourth line ( 0> X X _ _ _ _ _ _ _ _ <9, where an X
represents a rectangle) indicates that cartridges are installed in slots 0 and 1.
______________________
Note _______________________
There are no switches for setting a mechanical SCSI ID for the
tape drives. The SCSI IDs default to five. The MiniLibrary sets
the electronic SCSI ID very quickly, before any device can probe
the MiniLibrary, so the lack of a mechanical SCSI ID does not
cause any problems on the SCSI bus.
To set the SCSI ID, follow these steps:
1. From the Default Screen, press the Enter button to enter the Menu
Mode, displaying the Main Menu.
Configuring a Shared SCSI Bus for Tape Drive Use 8–53
____________________
Note _____________________
When you enter the Menu Mode, the Ready light goes out,
an indication that the module is off line, and all medium
changer commands from the host return a SCSI "not ready"
status until you exit the Menu Mode and the Ready light
comes on once again.
2. Depress the down arrow button until the Configure Menu item is
selected, then press the Enter button to display the Configure submenu.
____________________
Note _____________________
The control panel up and down arrows have an auto-repeat
feature. When you press either button for more than one-half
second, the control panel behaves as if you were pressing the
button about four times per second. The effect stops when
you release the button.
3. Press the down arrow button until the Set SCSI item is selected and
press the Enter button.
4. Select the tape drive (DLT0 Bus ID: or DLT1 Bus ID:) or library robotics
(LIB Bus ID:) for which you wish to change the SCSI bus ID. The default
SCSI IDs are as follows:
•
•
•
Lib Bus ID: 0
DLT0 Bus ID: 4
DLT1 Bus ID: 5
Use the up or down arrow button to select the item for which you need
to change the SCSI ID. Press the Enter button.
5. Use the up or down arrow button to scroll through the possible SCSI ID
settings. Press the Enter button when the desired SCSI ID is displayed.
6. Repeat steps 4 and 5 to set other SCSI bus IDs as necessary.
7. Press the Escape button repeatedly until the default menu is displayed.
8.11.2.1.2 Cabling the TL881 or TL891 DLT MiniLibrary
There are six 68-pin, high-density SCSI connectors on the back of the TL881
or TL891 DLT MiniLibrary standalone model or rackmount base unit. The
two leftmost connectors are for the library robotics controller. The middle
two are for tape drive 1. The two on the right are for tape drive 2 (if the
second tape drive is installed).
8–54 Configuring a Shared SCSI Bus for Tape Drive Use
______________________
Note _______________________
The tape drive SCSI connectors are labeled DLT1 (tape drive 1)
and DLT2 (tape drive 2). The control panel designation for the
drives is DLT0 (tape drive 1) and DLT1 (tape drive 2).
The default for the TL881 or TL891 DLT MiniLibrary is to place the robotics
controller and tape drive 1 on the same SCSI bus (Figure 8–17). A 0.3-meter
SCSI jumper cable is provided with the unit. Plug this cable into the second
connector (from the left) and the third connector. If the MiniLibrary has two
drives, place the second drive on the same SCSI bus with another 0.3-meter
SCSI bus jumper cable, or place it on its own SCSI bus.
______________________ Notes ______________________
The internal cabling of the TL881 and TL891 is too long to allow
external termination with a trilink/terminator combination.
Therefore, the TL881 or TL891 must be the last device on the
shared SCSI bus. They may not be removed from the shared
SCSI bus without stopping all ASE services that generate activity
on the bus.
To achieve system performance capabilities, we recommend
placing no more than two tape drives on a SCSI bus.
We recommend that tape devices be placed on separate shared
SCSI buses, and that there be no storage devices on the SCSI bus.
The cabling depends on whether or not there are one or two drives, and for
the two-drive configuration, if each drive is on a separate SCSI bus.
______________________
Note _______________________
It is assumed that the library robotics is on the same SCSI bus as
tape drive 1.
To connect the library robotics and one drive to a single shared SCSI bus,
follow these steps:
1. Connect a 328215-00X, BN21K, or BN21L between the last Y cable or
trilink connector on the bus to the leftmost connector (as viewed from
the rear) of the MiniLibrary. The 328215-004 is a 20-meter cable.
2. Install a 0.3-meter SCSI bus jumper between the rightmost robotics
connector (second connector from the left) and the left DLT1 connector
(the third connector from the left).
Configuring a Shared SCSI Bus for Tape Drive Use 8–55
3. Install an HD68 differential terminator (such as an H879-AA) on the
right DLT1 connector (the fourth connector from the left).
To connect the drive robotics and two drives to a single shared SCSI bus,
follow these steps:
1. Connect a 328215-00X, BN21K, or BN21L between the last trilink
connector on the bus to the leftmost connector (as viewed from the rear)
of the MiniLibrary.
2. Install a 0.3-meter SCSI bus jumper between the rightmost robotics
connector (the second connector from the left) and the left DLT1
connector (the third connector from the left).
3. Install a 0.3-meter SCSI bus jumper between the rightmost DLT1
connector (the fourth connector from the left) and the left DLT2
connector (the fifth connector from the left).
4. Install an HD68 differential (H879-AA) terminator on the right DLT2
connector (the rightmost connector).
To connect the drive robotics and one drive to one shared SCSI bus and the
second drive to a second shared SCSI bus, follow these steps:
1. Connect a 328215-00X, BN21K, or BN21L between the last trilink
connector on one shared SCSI bus to the leftmost connector (as viewed
from the rear) of the MiniLibrary.
2. Connect a 328215-00X, BN21K, or BN21L between the last trilink
connector on the second shared SCSI bus to the left DLT2 connector
(the fifth connector from the left).
3. Install a 0.3-meter SCSI bus jumper between the rightmost robotics
connector (the second connector from the left) and the left DLT1
connector (the third connector from the left).
4. Install an HD68 differential (H879-AA) terminator on the right DLT1
connector (the fourth connector from the left) and install another HD68
differential terminator on the right DLT2 connector (the rightmost
connector).
Figure 8–17 shows an example of a TruCluster configuration with a TL891
standalone MiniLibrary connected to two shared SCSI buses.
8–56 Configuring a Shared SCSI Bus for Tape Drive Use
Figure 8–17: TL891 Standalone Cluster Configuration
Network
Member
System
1
Member
System
2
Memory
Channel
Interface
Memory Channel
KZPBA-CB (ID 6)
Memory Channel
KZPBA-CB (ID 7)
KZPBA-CB (ID 7)
7
6
T
5
5
T
KZPBA-CB (ID 6)
T
1
1
T
Library
DLT1
6
Robotics
T
T
DS-DWZZH-03
7
2
3
4
T
DLT2
Controller B
HSZ70
Controller A
HSZ70
Expansion
Unit
Interface
StorageWorks
RAID Array 7000
TL891
0.3 m
SCSI Bus
Jumper
NOTE: This drawing is not to scale.
ZK-1627U-AI
Table 8–12 shows the components used to create the cluster shown in
Figure 8–17.
Table 8–12: Hardware Components Used to Create the Configuration
Shown in Figure 8–17
Description
Callout Number
BN38C or BN38D cablea
BN37A cableb
1
2
3
4
5
H8861-AA VHDCI trilink connector
H8863-AA VHDCI terminator
BN21W-0B Y cable
Configuring a Shared SCSI Bus for Tape Drive Use 8–57
Table 8–12: Hardware Components Used to Create the Configuration
Shown in Figure 8–17 (cont.)
Description
Callout Number
6
H879-AA terminator
328215-00X, BN21K, or BN21L cablec
7
a
b
c
The maximum length of the BN38C (or BN38D) cable on one SCSI bus segment must not exceed 25 meters.
The maximum length of the BN37A cable must not exceed 25 meters.
The maximum combined length of these cables must not exceed 25 meters.
8.11.2.2 Preparing a TL881 or TL891 Rackmount MiniLibrary for Shared SCSI
Bus Usage
A TL881 or TL891 MiniLibrary base unit may also be used in a rackmount
configuration with an expansion unit, data unit(s), and other base units, to
add tape drive and/or cartridge capacity to the configuration.
The expansion unit is installed above the TL881 or TL891 DLT MiniLibrary
base or data units in a SW500, SW800, or RETMA cabinet.
For complete hardware installation instructions, see the TL881 MiniLibrary
System User ’s Guide or TL891 MiniLibrary System User ’s Guide.
The topics in this section provide information on preparing the rackmount
TL881 or TL891 DLT MiniLibrary for use on a shared SCSI bus.
It is assumed that the expansion unit, base modules, and pass-through and
motor mechanism have been installed.
8.11.2.2.1 Cabling the Rackmount TL881 or TL891 DLT MiniLibrary
You must make the following connections to render the DLT MiniLibrary
system operational:
•
Expansion unit to the pass-through motor mechanism: The motor
mechanism cable is about 1 meter long and has a DB-15 connector
on each end. Connect it between the connector labeled Motor on the
expansion unit to the motor on the pass-through mechanism.
_____________________ Note _____________________
This cable is not shown in Figure 8–18 as the pass-through
mechanism is not shown in the figure.
•
Robotics control cables from the expansion unit to each base unit or
data unit: These cables have a DB-9 male connector on one end and
a DB-9 female connector on the other end. Connect the male end to
the Expansion Unit Interface connector on the base unit or Diagnostic
8–58 Configuring a Shared SCSI Bus for Tape Drive Use
connector on the data unit and the female end to any Expansion Modules
connector on the expansion unit.
_____________________ Note _____________________
It does not matter which interface connector a base unit or
data unit is connected to.
•
•
SCSI bus connection to the expansion unit robotics: Connect the shared
SCSI bus that will control the robotics to one of the SCSI connectors
on the expansion unit with a 328215-00X, BN21K, or BN21L cable.
Terminate the SCSI bus with an HD68 terminator (such as an H879-AA)
on the other expansion unit SCSI connector.
SCSI bus connection to each of the base module tape drives: Connect a
shared SCSI bus to one of the DLT1 or DLT2 SCSI connectors on each of
the base modules with 328215-00X, BN21K, or BN21L cables. Terminate
the other DLT1 or DLT2 SCSI bus connection with an HD68 terminator
(H879-AA).
You can daisy chain between DLT1 and DLT2 (if present) with a
0.3-meter SCSI bus jumper (supplied with the TL881 or TL891).
Terminate the SCSI bus at the tape drive on the end of the shared SCSI
bus with an HD68 terminator (H879-AA).
____________________
Notes
____________________
Do not connect a SCSI bus to the SCSI connectors for the
library connectors on the base modules.
We recommend that no more than two tape drives be on
a SCSI bus.
Figure 8–18 shows a TL891 DLT MiniLibrary configuration with an
expansion unit, a base units, and a data unit. The library robotics expansion
unit is on one shared SCSI bus and the two tape drives in the base unit are
on separate, shared SCSI buses. The data unit is not on a shared SCSI bus
as it does not contain any tape drives but tape cartridges only. Note that
the pass-through mechanism and cable to the library robotics motor is not
shown in this figure.
For more information on cabling the units, see Section 8.11.2.1.2. With the
exception of the robotics control on the expansion module, a rackmount
TL881 or TL891 DLT MiniLibrary is cabled in the same manner as a
tabletop unit.
Configuring a Shared SCSI Bus for Tape Drive Use 8–59
Figure 8–18: TL881 DLT MiniLibrary Rackmount Configuration
Network
Memory
Member System 1
Member System 2
Channel
Interface
Memory Channel
Memory Channel
T
6
7
5
KZPBA-CB (ID 6)
KZPBA-CB (ID 6)
KZPBA-CB (ID 6)
KZPBA-CB (ID 7)
T
KZPBA-CB (ID 7)
KZPBA-CB (ID 7)
5
5
T
T
7
7
1
1
T
DS-DWZZH-03
T
T
Diag
SCSI
Motor
6
2
3
4
T
Expansion
Unit
Controller B
HSZ70
Controller A
HSZ70
Expansion Modules
Library
Robotics
StorageWorks
DLT2
TL891
Robotics
Control
cables
RAID Array 7000
6
Base
Unit
0.3 Meter
Jumper
Cable
DLT1
Data
Unit
Diag
NOTE: This drawing is not to scale.
Robotic motor and pass through
mechanism not shown.
ZK-1628U-AI
Table 8–13 shows the components used to create the cluster shown in
Figure 8–18.
8–60 Configuring a Shared SCSI Bus for Tape Drive Use
Table 8–13: Hardware Components Used to Create the Configuration
Shown in Figure 8–18
Description
Callout Number
BN38C or BN38D cablea
BN37A cableb
1
2
3
4
5
6
H8861-AA VHDCI trilink connector
H8863-AA VHDCI terminator
BN21W-0B Y cable
H879-AA terminator
328215-00X, BN21K, or BN21L cablec
7
a
b
c
The maximum length of the BN38C (or BN38D) cable on one SCSI bus segment must not exceed 25 meters.
The maximum length of the BN37A cable must not exceed 25 meters.
The maximum combined length of these cables must not exceed 25 meters.
8.11.2.2.2 Configuring a Base Unit as a Slave to the Expansion Unit
The TL881/TL891 base units are shipped configured as standalone systems.
When they are used in conjunction with the MiniLibrary expansion unit, the
expansion unit must control the robotics of each of the base units. Therefore,
the base units must be configured as slaves to the expansion unit.
After the hardware and cables are installed, but before you power up
the expansion unit in a MiniLibrary system for the first time, you must
reconfigure each of the base units in the system as a slave. The expansion
unit will not have control over the base unit robotics when you power up the
MiniLibrary system, if you do not reconfigure the base unit as a slave.
To reconfigure a TL881/TL891 base unit as a slave to the MiniLibrary
expansion unit, perform the following procedure on each base unit in the
system:
1. Turn on the power switch on the TL881/TL891 base unit to be
reconfigured.
____________________
Note _____________________
Do not power on the expansion unit. Leave it powered off
until all base units have been reconfigured as slaves.
After a series of self-tests have executed, the default screen will be
displayed on the base module control panel:
Configuring a Shared SCSI Bus for Tape Drive Use 8–61
DLT0 Idle
DLT1 Idle
Loader Idle
0> _ _ _ _ _ _ _ _ _ _ <9
The default screen shows the state of the tape drives, loader, and
number of cartridges present for this base unit. A rectangle in place of
the underscore indicates that a cartridge is present in that location.
2. Press the Enter button to enter the Menu Mode, displaying the Main
Menu.
3. Depress the down arrow button until the Configure Menu item is
selected, then press the Enter button.
____________________
Note _____________________
The control panel up and down arrows have an auto-repeat
feature. When you press either button for more than one-half
second, the control panel behaves as if you were pressing the
button about four times per second. The effect stops when
you release the button.
4. Press the down arrow button until the Set Special Config menu is
selected and press the Enter button.
5. Press the down arrow button repeatedly until the Alternate Config item
is selected and press the Enter button.
6. Press the down arrow button to change the alternate configuration from
the default (Standalone) to Slave. Press the Enter button.
7. After the selection stops flashing and the control panel indicates that
the change is not effective until a reboot, press the Enter button.
8. When the Special Configuration menu reappears, turn the power
switch off and then on again to cycle the power. The base unit is now
reconfigured as a slave to the expansion unit.
9. Repeat the steps for each TL881/TL891 base unit present that is to
be a slave to the expansion unit.
8.11.2.2.3 Powering Up the TL881/TL891 DLT MiniLibrary
When turning on power to the TL881 or TL891 DLT MiniLibrary, power
must be applied to the expansion unit simultaneously or after power is
applied to the the base units and data units. If the expansion unit is powered
on first, its inventory of modules may be incorrect and the contents of some
or all of the modules will be inaccessible to the system and to the host.
8–62 Configuring a Shared SCSI Bus for Tape Drive Use
When the expansion unit comes up, it will communicate with each base and
data unit through the expansion unit interface and inventory the number
of base units, tape drives, data units, and cartridges present in each base
and data unit. After the MiniLibrary configuration has been determined, the
expansion unit will communicate with each base and data unit and indicate
to the modules which cartridge group that base or data unit contains.
When all initialization communication between the expansion module and
each base and data unit has completed, the base and data units will display
their cartridge numbers according to the remapped cartridge inventory.
8.11.2.2.4 Setting the SCSI IDs for a Rackmount TL881 or TL891 DLT MiniLibrary
After the base units have been reconfigured as slaves, each base unit control
panel still provides tape drive status and error information, but all control
functions are carried out from the expansion unit control panel. This
includes setting the SCSI ID for each of the tape drives present.
To set the SCSI IDs for the tape drives in a TL881 or TL891 DLT MiniLibrary
rackmount configuration, follow these steps:
1. Apply power to the MiniLibrary, ensuring that you power up the
expansion unit after or at the same time as the base and data units.
2. Wait until power-on self-tests (POST) have terminated and the
expansion unit and each base and data unit display the default screen.
3. At the expansion unit control panel, press the Enter button to display
the Main Menu.
4. Press the down arrow button until the Configure Menu item is selected,
and then press the Enter button to display the Configure submenu.
5. Press the down arrow button until the Set SCSI item is selected and
press the Enter button.
6. Press the up or down arrow button to select the appropriate tape drive
(DLT0 Bus ID:, DLT1 Bus ID:, DLT2 Bus ID:, and so on) or library
robotics (Library Bus ID:) for which you wish to change the SCSI bus
ID. In a configuration with three base units, and assuming that each
base unit has two tape drives, the top base unit contains DLT0 and
DLT1. The next base unit down contains DLT2 and DLT3. The next
base unit contains DLT4 and DLT5. The default SCSI IDs, after being
reconfigured by the expansion unit, are as follows:
•
•
•
•
Library Bus ID: 0
DLT0 Bus ID: 1
DLT1 Bus ID: 2
DLT2 Bus ID: 3
Configuring a Shared SCSI Bus for Tape Drive Use 8–63
•
•
•
DLT3 Bus ID: 4
DLT4 Bus ID: 5
DLT5 Bus ID: 6
7. Press Enter when you have the item selected for which you wish to
change the SCSI ID.
8. Use the up and down arrows to select the desired SCSI ID. Press the
Enter button to save the new selection.
9. Press the Escape button once to return to the Set SCSI Submenu to
select another tape drive or the library robotics, and then repeat steps 6,
7, and 8 to set the SCSI ID.
10. If there are other items you wish to configure, press the Escape button
until the Configure submenu is displayed, then select the item to be
configured. Repeat this procedure for each item you wish to configure.
11. If there are no more items to be configured, press the Escape button
until the Default window is displayed.
______________________
Note _______________________
You do not have to cycle power to set the SCSI IDs.
8.12 Compaq ESL9326D Enterprise Library
The topics in this section provide an overview and hardware configuration
information on preparing the ESL9326D Enterprise Library for use on a
shared SCSI bus with the TruCluster Server.
8.12.1 General Overview
The Compaq StorageWorks ESL9326D Enterprise Library is the first
building block of the Compaq ESL 9000 series tape library.
For more information on the ESL9326D Enterprise Library, see the following
Compaq StorageWorks ESL9000 Series Tape Library documentation:
•
•
•
•
Unpacking and Installation Guide (146585-001)
Reference Guide (146583-001)
Maintenance and Service Guide (155898-001)
Tape Drive Upgrade Guide (146582-001)
8–64 Configuring a Shared SCSI Bus for Tape Drive Use
______________________
Note _______________________
These tape devices have been qualified for use on shared SCSI
buses with both the KZPSA-BB and KZPBA-CB host bus
adapters.
8.12.2 ESL9326D Enterprise Library Overview
The ESL9326D Enterprise Library is an enterprise Digital Linear Tape
(DLT) automated tape library with from 6 to 16 fast-wide, differential tape
drives. This tape library uses the 35/70 DLT (DS-TZ89N-AV) differential
tape drives. The SCSI bus connectors are 68-pin, high-density.
The ESL9326D Enterprise Library has a capacity of 326 DLT cartridges in a
fixed storage array (back wall, inside the left door, and inside the right door).
This provides a storage capacity of 11.4 TB uncompressed for the ESL9326D
Enterprise Library using DLT Tape IV cartridges. The library can also use
DLT Tape III or IIIXT tape cartridges.
The ESL9326D Enterprise Library is available as six different part numbers,
based on the number of tape drives:
Number of Tape Drives
Order Number
146205-B23
146205-B24
146205-B25
146205-B26
146205-B27
146205-B28
6
8
10
12
14
16
A tape library with a capacity for additional tape drives may be upgraded
with part number 146209-B21, which adds a 35/70 DLT tape drive. See the
Compaq StorageWorks ESL9000 Series Tape Library Tape Drive Upgrade
Guide (146582-001) for more information.
8.12.3 Preparing the ESL9326D Enterprise Library for Shared SCSI
Bus Usage
The ESL9326D Enterprise Library contains library electronics (robotic
controller) and from 6 to 16 35/70 DLT (DS-TZ89N-AV) fast-wide, differential
Digital Linear Tape (DLT) tape drives.
Tape devices are supported only on those shared SCSI buses that use the
KZPSA-BB or KZPBA-CB host bus adapters.
Configuring a Shared SCSI Bus for Tape Drive Use 8–65
______________________ Notes ______________________
The ESL9326D Enterprise Library is cabled internally for two
35/70 DLT tape drives on each SCSI bus. It arrives with the
library electronics cabled to tape drives 0 and 1. Every other
pair of tape drives is cabled together (2 and 3, 4 and 5, 6 and
7, and so on).
An extra SCSI bus jumper cable is provided with the ESL9326D
Enterprise Library for those customers that are short on SCSI
buses, and want to jumper two SCSI buses together and place
four tape drives on the same SCSI bus.
We recommend that you place no more that two 35/70 DLT tape
drives on a shared SCSI bus.
We also recommended that storage not be placed on shared SCSI
buses that have tape drives.
The following sections describe how to prepare the ESL9326D Enterprise
Library in more detail.
8.12.3.1 ESL9326D Enterprise Library Robotic and Tape Drive Required Firmware
Library electronics firmware V1.22 is the minimum firmware version that
supports TruCluster Server.
The 35/70 DLT tape drives require V80 or later firmware.
8.12.3.2 Library Electronics and Tape Drive SCSI IDs
The default robotics and tape drive SCSI IDs are shown in Figure 8–19. If
these SCSI IDs are not acceptable for your configuration and you need to
change them, follow the steps in the Compaq StorageWorks ESL9000 Series
Tape Library Reference Guide (146583-001).
8.12.3.3 ESL9326D Enterprise Library Internal Cabling
The default internal cabling for the ESL9326D Enterprise Library is to place
two 35/70 DLT tape drives on one SCSI bus.
Figure 8–19 shows the default cabling for an ESL9326D Enterprise Library
with 16 tape drives. Note that each pair of tape drives is cabled together
internally to place two drives on a single SCSI bus. If your model has fewer
drives, all internal cabling is supplied. The terminators for the drives not
present are not installed on the SCSI bulkhead.
8–66 Configuring a Shared SCSI Bus for Tape Drive Use
Figure 8–19: ESL9326D Internal Cabling
Tape Drive 8
SCSI ID 2
Tape Drive 0
SCSI ID 2
Tape Drive 1
SCSI ID 3
Tape Drive 9
SCSI ID 3
Tape Drive 2
SCSI ID 4
Tape Drive 10
SCSI ID 4
Tape Drive 11
SCSI ID 5
Tape Drive 3
SCSI ID 5
Tape Drive 12
SCSI ID 2
Tape Drive 4
SCSI ID 2
Tape Drive 13
SCSI ID 3
Tape Drive 5
SCSI ID 3
Tape Drive 14
SCSI ID 4
Tape Drive 6
SCSI ID 4
Tape Drive 15
SCSI ID 5
Tape Drive 7
SCSI ID 5
Robotics
SCSI ID 0
SCSI Bulkhead
P
O
N
M
L
K
J
I
A
B
C
D
E
F
G
H
Q
R
T
T
T
T
T
T
T
T
SCSI
Bus In
ZK-1705U-AI
______________________
Note _______________________
Each internal cable is up to 2.5 meters long. The length of
the internal cables, two per SCSI bus, must be taken into
consideration when ordering SCSI bus cables.
The maximum length of a differential SCSI bus segment is 25
meters, and the internal tape drive SCSI bus length is 5 meters.
Therefore, you must limit the external SCSI bus cables to 20
meters maximum.
Configuring a Shared SCSI Bus for Tape Drive Use 8–67
8.12.3.4 Connecting the ESL9326D Enterprise Library to the Shared SCSI Bus
The ESL9326D Enterprise Library has 5 meters of internal SCSI bus cabling
for each pair of tape drives. Because of the internal SCSI bus lengths, it
is not possible to use a trilink connector or Y cable to terminate the SCSI
bus external to the tape library as is done with other devices on the shared
SCSI bus. Each SCSI bus must be terminated at the end of the SCSI bus by
installing a terminator on the SCSI bulkhead SCSI connector. Therefore,
TruCluster Server configurations using the ESL9326D Enterprise Library
must ensure that the tape library is on the end of the shared SCSI bus.
______________________
Note _______________________
We recommend that disk storage devices be placed on separate
shared SCSI buses.
Use 328215-001 (5-meter), 328215-002 (10-meter), 328215-003 (15-meter),
328215-004 (20-meter), or BN21K (BN21L) cables of the appropriate length
to connect the ESL9326D Enterprise Library to a shared SCSI bus. Do
not use a cable longer than 20 meters. Terminate each SCSI bus with
a 330563-001 (or H879-AA) HD-68 terminator. Connect the cables and
terminator on the SCSI bulkhead SCSI connectors as shown in Table 8–14
to form shared SCSI buses.
Table 8–14: Shared SCSI Bus Cable and Terminator Connections for the
ESL9326D Enterprise Library
Install HD68 Terminator
on Connector:
Tape Drives on Shared
SCSI Bus
Connect SCSI Cable
to Connector:
0, 1, and library electronicsa
Q
C
E
B
D
F
2, 3
4, 5
6, 7
G
I
H
J
8, 9
10, 11
12, 13
K
M
O
L
N
P
14, 15
a
Install .3-meter jumper cable part number 330582-001 between SCSI connectors R and A to place the
library electronics on the SCSI bus with tape drives 0 and 1.
8–68 Configuring a Shared SCSI Bus for Tape Drive Use
______________________ Notes ______________________
Each ESL9326D Enterprise Library arrives with one 330563-001
HD68 terminator for each pair of tape drives (one SCSI bus). The
kit also includes at least one 330582-001 jumper cable to connect
the library electronics to tape drives 0 and 1.
Tape libraries with more than six tape drives include extra
330582-01 jumper cables in case the customer is short on host bus
adapters and wants to place more than two tape drives on a single
SCSI bus (a configuration that we do not recommend).
Configuring a Shared SCSI Bus for Tape Drive Use 8–69
9
Configurations Using External
Termination or Radial Connections
to Non-UltraSCSI Devices
This chapter describes the requirements for the shared SCSI bus using:
•
•
Externally terminated TruCluster Server configurations
Radial configurations with non-UltraSCSI RAID array controllers
In addition to using only the supported hardware, adhering to the
requirements described in this chapter will ensure that your cluster operates
correctly.
This chapter discusses the following topics:
•
•
Using SCSI bus signal converters (Section 9.1)
SCSI bus termination in externally terminated TruCluster Server
configurations (Section 9.2)
•
•
Overview of the BA350, BA356, and UltraSCSI BA356 disk storage
shelves (Section 9.3)
Preparing the storage configuration for external termination using Y
cables and trilinks (Section 9.4)
–
–
–
Preparing the storage shelves for an externally terminated
TruCluster Server configuration (Section 9.4.1)
Connecting multiple storage shelves, for instance a BA350 and a
BA356, two BA356s, or two UltraSCSI BA356s (Section 9.4.2)
Using the HSZ20, HSZ40, or HSZ50 RAID array controllers
(Section 9.4.3)
•
Radial configurations using the HSZ40 or HSZ50 RAID array controllers
(Section 9.4.4)
Introductory information covering SCSI bus configuration concepts (SCSI
bus speed, data path, and so on) and SCSI bus configuration requirements
can be found in Chapter 3.
Configurations Using External Termination or Radial Connections to
Non-UltraSCSI Devices 9–1
9.1 Using SCSI Bus Signal Converters
A SCSI bus signal converter allows you to couple a differential bus segment
to a single-ended bus segment, allowing the mixing of differential and
single-ended devices on the same SCSI bus to isolate bus segments for
maintenance purposes.
Each SCSI signal converter has a single-ended side and a differential side as
follows:
•
•
•
DWZZA — 8-bit data path
DWZZB — 16-bit data path
DS-BA35X-DA 16-bit personality module
______________________
Note _______________________
Some UltraSCSI documentation uses the UltraSCSI "bus
expander" term when referring to the DWZZB and UltraSCSI
signal converters. Other UltraSCSI documentation refers to some
UltraSCSI products as bus extender/converters.
For TruCluster Server there are no supported standalone
UltraSCSI bus expanders (DWZZC).
In this manual, any device that converts a differential signal to
a single-ended signal is referred to as a signal converter (the
DS-BA35X-DA personality module contains a DWZZA-on-a-chip
or DOC chip).
A SCSI signal converter is required when you want to connect devices with
different transmission modes.
9.1.1 Types of SCSI Bus Signal Converters
Signal converters can be standalone units or StorageWorks building blocks
(SBBs) that are installed in a storage shelf disk slot. You must use the signal
converter module that is appropriate for your hardware configuration.
For example, use a DWZZA-VA signal converter to connect a wide,
differential host bus adapter to a BA350 (single-ended and narrow) storage
shelf, but use a DWZZB-VW signal converter to connect a wide, differential
host bus adapter to a non-UltraSCSI BA356 (single-ended and wide) storage
shelf. The DS-BA35X-DA personality module is used in an UltraSCSI BA356
to connect an UltraSCSI host bus adapter to the single-ended disks in the
UltraSCSI BA356. You could install a DWZZB-VW in an UltraSCSI BA356,
9–2 Configurations Using External Termination or Radial Connections to Non-UltraSCSI Devices
but you would waste a disk slot and it would not work with a KZPBA-CB if
there are any UltraSCSI disks in the storage shelves.
The following sections discuss the DWZZA and DWZZB signal converters
and the DS-BA35X-DA personality module.
9.1.2 Using the SCSI Bus Signal Converters
The DWZZA and DWZZB signal converters are used in the BA350 and BA356
storage shelves. They have removable termination. The DS-BA35X-DA
personality module is used in the UltraSCSI BA356. It has switch selectable
differential termination. The single-ended termination is active termination.
The following sections describe termination for these signal converters in
more detail.
9.1.2.1 DWZZA and DWZZB Signal Converter Termination
Both the single-ended side and the differential side of each DWZZA and
DWZZB signal converter has removable termination. To use a signal
converter, you must remove the termination in the differential side
and attach a trilink connector to this side. To remove the differential
termination, remove the five 14-pin termination resistor SIPs (located near
the differential end of the signal converter). You can attach a terminator to
the trilink connector to terminate the differential bus. If you detach the
trilink connector from the signal converter, the shared SCSI bus is still
terminated (provided there is termination power).
You must keep the termination in the single-ended side to provide
termination for one end of the BA350 or BA356 single-ended SCSI bus
segment. Verify that the termination is active. A DWZZA should have
jumper J 2 installed. J umpers W1 and W2 should be installed in a DWZZB.
Figure 9–1 shows the status of internal termination for a standalone SCSI
signal converter that has a trilink connector attached to the differential side.
Configurations Using External Termination or Radial Connections to
Non-UltraSCSI Devices 9–3
Figure 9–1: Standalone SCSI Signal Converter
T
T
Single-ended
side
Differential side with trilink
attached
ZK-1050U-AI
Figure 9–2 shows the status of internal termination for an SBB SCSI signal
converter that has a trilink connector attached to the differential side.
Figure 9–2: SBB SCSI Signal Converter
T
T
Single-ended
side
Differential side with trilink
attached
ZK-1576U-AI
9.1.2.2 DS-BA35X-DA Termination
The UltraSCSI BA356 shelf uses a 16-bit differential UltraSCSI personality
module (DS-BA35X-DA) as the interface between the UltraSCSI differential
bus and the UltraSCSI single-ended bus in the UltraSCSI BA356.
The personality module controls termination for the external differential
UltraSCSI bus segment, and for both ends of the internal single-ended bus
segment.
For normal cluster operation, the differential termination must be disabled
since a trilink connector will be installed on personality module connector
J A1, allowing the use of the UltraSCSI BA356 (or two UltraSCSI BA356s)
in the middle of the bus or external termination for an UltraSCSI BA356
on the end of the bus.
Switch pack 4 switches S4-1 and S4-2 are set to ON to disable the personality
module differential termination. The switches have no effect on the BA356
internal, single-ended UltraSCSI bus termination.
9–4 Configurations Using External Termination or Radial Connections to Non-UltraSCSI Devices
______________________ Notes ______________________
S4-3 and S4-4 have no function on the DS-BA35X-DA personality
module.
See Section 9.3.2.2 for information on how to select the device
SCSI IDs in an UltraSCSI BA356.
Figure 9–3 shows the relative positions of the two DS-BA35X-DA switch
packs.
Figure 9–3: DS-BA35X-DA Personality Module Switches
OFF ON
1
2
3
4
SCSI Bus
Termination
Switch S4
ON
OFF
1 2 3 4 5 6 7
SCSI Bus Address
Switch S3
ZK-1411U-AI
9.2 Terminating the Shared SCSI Bus
You must properly connect devices to a shared SCSI bus. In addition, you
can terminate only the beginning and end of each SCSI bus segment (either
single-ended or differential).
There are two rules for SCSI bus termination:
•
•
There are only two terminators for each SCSI bus segment.
If you do not use an UltraSCSI hub, bus termination must be external.
Note that you may use external termination with an UltraSCSI hub,
but is is not the recommended way.
Configurations Using External Termination or Radial Connections to
Non-UltraSCSI Devices 9–5
Whenever possible, connect devices to a shared bus so that they can be
isolated from the bus. This allows you to disconnect devices from the bus
for maintenance purposes without affecting bus termination and cluster
operation. You also can set up a shared SCSI bus so that you can connect
additional devices at a later time without affecting bus termination.
______________________ Notes ______________________
With the exception of the TZ885, TZ887, TL890, TL891, and
TL892, tape devices can only be installed at the end of a shared
SCSI bus. These tape devices are the only supported tape devices
that can be terminated externally.
We recommend that tape loaders be on a separate shared SCSI
bus to allow normal shared SCSI bus termination for those shared
SCSI buses without tape loaders.
Most devices have internal termination. For example, the KZPSA and
KZPBA host bus adapters, BA350 and BA356 storage shelves, and the
DWZZA and DWZZB SCSI bus signal converters have internal termination.
Depending on how you set up a shared bus, you may have to enable or
disable device termination.
Unless you are using an UltraSCSI hub, if you use a device’s internal
termination to terminate a shared bus, and you disconnect the bus cable
from the device, the bus will not be terminated and cluster operation will
be impaired. Therefore, unless you use an UltraSCSI hub, you must use
external termination, enabling you to detach the device without affecting bus
termination. The use of UltraSCSI hubs with UltraSCSI devices is discussed
in Section 3.5 and Section 3.6. The use of a DS-DWZZH-03 UltraSCSI hub
with externally terminated host bus adapters is discussed in Section 9.4.3.
To be able to externally terminate a bus and connect and disconnect devices
without affecting bus termination, remove the device termination and use Y
cables or trilink connectors to connect a device to a shared SCSI bus.
By attaching a Y cable or trilink connector to an unterminated device, you
can locate the device in the middle or at the end of the shared bus. If
the device is at the end of a bus, attach an H879-AA terminator to the
BN21W-0B Y cable or H885-AA trilink connector to terminate the bus. For
UltraSCSI devices, attach an H8863-AA terminator to the H8861 trilink
connector. If you disconnect the Y cable or trilink connector from the device,
the shared bus is still terminated and the shared SCSI bus is still operable.
In addition, you can attach a Y cable or a trilink connector to a properly
terminated shared bus without connecting the Y cable or trilink connector
to a device. If you do this, you can connect a device to the Y cable or trilink
9–6 Configurations Using External Termination or Radial Connections to Non-UltraSCSI Devices
connector at a later time without affecting bus termination. This allows you
to expand your configuration without shutting down the cluster.
Figure 9–4 shows a BN21W-0B Y cable, which you may attach to a
KZPSA-BB or KZPBA-CB SCSI adapter that has had its onboard termination
removed. You can also use the BN21W-0B Y cable with a HSZ40 or HSZ50
controller or the unterminated differential side of a SCSI signal converter.
______________________
Note _______________________
You will normally use a Y cable on a KZPSA-BB or KZPBA-CB
host bus adapter where there is not room for an H885-AA trilink,
and a trilink connector elsewhere.
Figure 9–4: BN21W-0B Y Cable
Figure 9–5 shows an HD68 trilink connector (H885-AA), which you
may attach to a KZPSA-BB or KZPBA-CB adapter that has its onboard
termination removed, an HSZ40 or HSZ50 controller, or the unterminated
differential side of a SCSI signal converter.
Configurations Using External Termination or Radial Connections to
Non-UltraSCSI Devices 9–7
Figure 9–5: HD68 Trilink Connector (H885-AA)
REAR VIEW
FRONT VIEW
ZK-1140U-AI
______________________
Note _______________________
If you connect a trilink connector to a SCSI bus adapter, you
may block access to an adjacent PCI slot. If this occurs, use a Y
cable instead of the trilink connector. This is the case with the
KZPBA-CB and KZPSA-BB SCSI adapters on some AlphaServer
systems.
Use the H879-AA terminator to terminate one leg of a BN21W-0B Y cable
or H885-AA trilink.
Use an H8861-AA VHDCI trilink connector (see Figure 3–1) with a
DS-BA35X-DA personality module to daisy chain two UltraSCSI BA356s
or to terminate external to the UltraSCSI BA356 storage shelf. Use the
H8863-AA VHDCI terminator with the H8861-AA trilink connector.
9.3 Overview of Disk Storage Shelves
The following sections provide an introduction to the BA350, BA356, and
UltraSCSI BA356 disk storage shelves.
9–8 Configurations Using External Termination or Radial Connections to Non-UltraSCSI Devices
9.3.1 BA350 Storage Shelf
Up to seven narrow (8-bit) single-ended StorageWorks building blocks
(SBBs) can be installed in the BA350. Their SCSI IDs are based upon the
slot they are installed in. For instance, a disk installed in BA350 slot 0 has
SCSI ID 0, a disk installed in BA350 slot 1 has SCSI ID 1, and so forth.
______________________
Note _______________________
Do not install disks in the slots corresponding to the host SCSI
IDs (usually SCSI ID 6 and 7 for a two-node cluster).
You use a DWZZA-VA as the interface between the wide, differential shared
SCSI bus and the BA350 narrow, single-ended SCSI bus segment.
______________________
Note _______________________
Do not use a DWZZB-VW in a BA350. The use of the wide
DWZZB-VW on the narrow single-ended bus will result in
unterminated data lines in the DWZZB-VW, which will cause
SCSI bus errors.
The BA350 storage shelf contains internal SCSI bus termination and a SCSI
bus jumper. The jumper is not removed during normal operation.
The BA350 can be set up for two-bus operation, but that option is not very
useful for a shared SCSI bus and is not covered in this manual.
Figure 9–6 shows the relative locations of the BA350 SCSI bus terminator
and SCSI bus jumper. They are accessed from the rear of the box. For
operation within a TruCluster Server cluster, both the J jumper and T
terminator must be installed.
Configurations Using External Termination or Radial Connections to
Non-UltraSCSI Devices 9–9
Figure 9–6: BA350 Internal SCSI Bus
JA1
JB1
0
T
1
2
3
4
J
5
6
POWER (7)
ZK-1338U-AI
9.3.2 BA356 Storage Shelf
There are two variations of the BA356 used in TruCluster Server clusters:
the BA356 (non-UltraSCSI BA356) and the UltraSCSI BA356.
An example of the non-UltraSCSI BA356 is the BA356-KC, which has a
wide, single-ended internal SCSI bus. It has a BA35X-MH 16-bit personality
module (only used for SCSI ID selection) and a 150-watt power supply. It is
referred to as the non-UltraSCSI BA356 or BA356 in this manual. You use a
DWZZB-VW as the interface between the wide, differential shared SCSI bus
and the BA356 wide, single-ended SCSI bus segment.
9.3.2.1 Non-UltraSCSI BA356 Storage Shelf
The non-UltraSCSI BA356, like the BA350, can hold up to seven
StorageWorks building blocks (SBBs). However, unlike the BA350, these
SBBs are wide devices and can therefore support up to 16 disks (in two
BA356 shelves). Also, like the BA350, the SBB SCSI IDs are based upon
the slot they are installed in. The switches on the personality module
(BA35X-MH) determine whether the disks respond to SCSI IDs 0 through 6
(slot 7 is the power supply) or 8 through 14 (slot 15 is the power supply). To
9–10 Configurations Using External Termination or Radial Connections to Non-UltraSCSI Devices
select SCSI IDs 0 through 6, set the personality module address switches 1
through 7 to off. To select SCSI IDs 8 through 14, set personality module
address switches 1 through 3 to on and switches 4 through 7 to off.
Figure 9–7 shows the relative location of the BA356 SCSI bus jumper,
BA35X-MF. The jumper is accessed from the rear of the box. For operation
within a TruCluster Server cluster, you must install the J jumper in the
normal position, behind slot 6. Note that the SCSI bus jumper is not in the
same position in the BA356 as in the BA350.
Termination for the BA356 single-ended bus is on the personality module,
and is active unless a cable is installed on J B1 to daisy chain the single-ended
SCSI bus in two BA356 storage shelves together. In this case, when the
cable is connected to J B1, the personality module terminator is disabled.
Daisy chaining the single-ended bus between two BA356s is not used in
clusters. We use DWZZB-VWs (with an attached H885-AA trilink connector)
in each BA356 to connect the wide-differential connection from the host
adapters to both BA356s in parallel. The switches on the personality module
of one BA356 are set for SCSI IDs 0 through 7 and the switches on the
personality module of the other BA356 are set for SCSI IDs 8 through 14.
______________________
Note _______________________
Do not install a narrow disk in a BA356 that is enabled for SCSI
IDs 8 through 14. The SCSI bus will not operate correctly because
the narrow disks cannot recognize wide addresses.
Like the BA350, you can set up the BA356 for two-bus operation by installing
a SCSI bus terminator (BA35X-ME) in place of the SCSI bus jumper.
However, like the BA350, two-bus operation in the BA356 is not very useful
for a TruCluster Server cluster.
You can use the position behind slot 1 in the BA356 to store the SCSI bus
terminator or jumper.
Figure 9–7 shows the relative locations of the BA356 SCSI bus jumper and
the position for storing the SCSI bus jumper, if you do install the terminator.
For operation within a TruCluster Server cluster, you must install the J
jumper.
Configurations Using External Termination or Radial Connections to
Non-UltraSCSI Devices 9–11
Figure 9–7: BA356 Internal SCSI Bus
JA1
JB1
0
1
2
3
4
5
6
J
POWER (7)
ZK-1339U-AI
Note that J A1 and J B1 are located on the personality module (in the top of
the box when it is standing vertically). J B1, on the front of the module, is
visible. J A1 is on the left side of the personality module as you face the front
of the BA356, and is hidden from the normal view.
To determine if a jumper module or terminator module is installed in a
BA356, remove the devices from slots 1 and 6 and note the following pin
locations (see Figure 9–8):
•
The identification pin on a jumper module aligns with the top hole in
the backplane.
•
The identification pin on a terminator module aligns with the bottom
hole in the backplane.
9–12 Configurations Using External Termination or Radial Connections to Non-UltraSCSI Devices
Figure 9–8: BA356 Jumper and Terminator Module Identification Pins
Slot 6
Jumper
Pin
Slot 1
Jumper
Pin
Slot 6
Terminator
Pin
Slot 1
Terminator
Pin
ZK-1529U-AI
9.3.2.2 UltraSCSI BA356 Storage Shelf
The UltraSCSI BA356 (DS-BA356-J F or DS-BA356-KH) has a single-ended,
wide UltraSCSI bus. The DS-BA35X-DA personality module provides the
interface between the internal, single-ended UltraSCSI bus segment and the
shared, wide, differential UltraSCSI bus. The UltraSCSI BA356 uses a
180-watt power supply.
An older, non-UltraSCSI BA356 that has been retrofitted with a BA35X-HH
180-watt power supply and DS-BA35X-DA personality module is still only
FCC certified for Fast 10 configurations (see Section 3.2.4 for a discussion on
bus speed).
The UltraSCSI BA356 can hold up to seven StorageWorks building blocks
(SBBs). These SBBs are UltraSCSI single-ended wide devices. The disk
SCSI IDs are based upon the slot they are installed in. The S3 switches
on the personality module (DS-BA35X-DA) determine whether the disks
respond to SCSI IDs 0 through 6 (slot 7 is the power supply) or 8 through 14
(slot 15 is the power supply). To select SCSI IDs 0 through 6, set switches
S3-1 through S3-7 to off. To select SCSI IDs 8 through 14, set personality
module address switches S3-1 through S3-3 to on and switches S3-4 through
S3-7 to off.
The jumper module is positioned behind slot 6 as with the non-UltraSCSI
BA356 shown in Figure 9–7. For operation within a TruCluster Server
cluster, you must install the J jumper. You verify the presence or absence
of the jumper or terminator modules the same as for the non-UltraSCSI
Configurations Using External Termination or Radial Connections to
Non-UltraSCSI Devices 9–13
BA356, as shown in Figure 9–8. With proper lighting you will be able to see
a J or T near the hole where the pin sticks through.
Termination for both ends of the UltraSCSI BA356 internal, single-ended
bus is on the personality module, and is always active. Termination for
the differential UltraSCSI bus is also on the personality module, and
is controlled by the SCSI bus termination switches, switch pack S4.
DS-BA35X-DA termination is discussed in Section 9.1.2.2.
9.4 Preparing the Storage for Configurations Using
External Termination
A TruCluster Server cluster provides you with high data availability through
the cluster file system (CFS), the device request dispatcher (DRD), service
failover through the cluster application availability (CAA) subsystem,
disk mirroring, and fast file system recovery. TruCluster Server supports
mirroring of the clusterwide root (/) file system, the member-specific boot
disks, and the cluster quorum disk through hardware RAID only. You can
mirror the clusterwide /usr and /var file systems and the data disks using
the Logical Storage Manager (LSM) technology. You must determine the
storage configuration that will meet your needs. Mirroring disks across two
shared buses provides the most highly available data.
Disk devices used on the shared bus must be located in a supported storage
shelf. Before you connect a storage shelf to a shared SCSI bus, you must
install the disks in the unit. Before connecting a RAID array controller
to a shared SCSI bus, install the disks and configure the storagesets. For
detailed information about installation and configuration, see your storage
shelf (or RAID array controller) documentation.
After completing the following sections and setting up your RAID
storagesets, you should be ready to cable your host bus adapters to storage
when they have been installed (see Chapter 10).
The following sections describe how to prepare storage for a shared SCSI bus
and external termination for:
•
•
•
•
A BA350, a BA356, and an UltraSCSI BA356
Two BA356s
Two UltraSCSI BA356s
An HSZ20, HSZ40, or HSZ50 RAID array controller
If you need to use a BA350 or non-UltraSCSI BA356 with an UltraSCSI
BA356 storage shelve, extrapolate the needed information from Section 9.4.1
and Section 9.4.2.
9–14 Configurations Using External Termination or Radial Connections to Non-UltraSCSI Devices
Later sections describe how to install cables to configure an HSZ20, HSZ40,
or HSZ50 in a TruCluster Server configuration with two member systems.
9.4.1 Preparing BA350, BA356, and UltraSCSI BA356 Storage Shelves
for an Externally Terminated TruCluster Server Configuration
You may be using the BA350, BA356, or UltraSCSI BA356 storage shelves in
your TruCluster Server configuration as follows:
•
A BA350 storage shelf provides access to SCSI devices through an
8-bit, single-ended, and narrow SCSI-2 interface. It can be used with a
DWZZA-VA and connected to a differential shared SCSI bus.
•
A non-Ultra BA356 storage shelf provides access to SCSI devices
through a 16-bit, single-ended, and wide SCSI-2 interface. In a cluster
configuration, you would connect a non-Ultra BA356 to the shared SCSI
bus using DWZZB-VW.
•
An UltraSCSI BA356 storage shelf provides access to UltraSCSI devices
through a 16-bit, single-ended, wide UltraSCSI interface. In a cluster
configuration, you would connect an UltraSCSI BA356 to the shared
SCSI bus through the DS-BA35X-DA personality module.
The following sections discuss the steps necessary to prepare the individual
storage shelves, and then connect two storage shelves together to provide
the additional storage.
______________________
Note _______________________
This material has been written with the premise that there are
only two member systems in any TruCluster Server configuration
using direct connect disks for storage. Using this assumption,
and further assuming that the member systems use SCSI IDs 6
and 7, the storage shelf housing disks in the range of SCSI IDs 0
through 6 can only use SCSI IDs 0 through 5.
If there are more than two member systems, additional disk slots
will be needed to provide the additional member system SCSI IDs.
9.4.1.1 Preparing a BA350 Storage Shelf for Shared SCSI Usage
To prepare a BA350 storage shelf for usage on a shared SCSI bus, follow
these steps:
1. Ensure that the BA350 storage shelf’s internal termination and jumper
is installed (see Section 9.3.1 and Figure 9–6).
Configurations Using External Termination or Radial Connections to
Non-UltraSCSI Devices 9–15
2. You will need a DWZZA-VA signal converter for the BA350. Ensure
that the DWZZA-VA single-ended termination jumper, J 2, is installed.
Remove the termination from the differential end by removing the five
14-pin differential terminator resistor SIPs.
3. Attach an H885-AA trilink connector to the DWZZA-VA 68-pin
high-density connector.
4. Install the DWZZA-VA in slot 0 of the BA350.
9.4.1.2 Preparing a BA356 Storage Shelf for Shared SCSI Usage
To prepare a BA356 storage shelf for shared SCSI bus usage, follow these
steps:
1. You need either a DWZZB-AA or DWZZB-VW signal converter.
The DWZZB-VW is more commonly used. Verify signal converter
termination as follows:
•
Ensure that the DWZZB W1 and W2 jumpers are installed to enable
the single-ended termination at one end of the bus. The other end of
the BA356 single-ended SCSI bus is terminated on the personality
module.
•
Remove the termination from the differential side of the DWZZB by
removing the five 14-pin differential terminator resistor SIPs. The
differential SCSI bus will be terminated external to the DWZZB.
2. Attach an H885-AA trilink connector to the DWZZB 68-pin high-density
connector.
3. Set the switches on the BA356 personality module as follows:
•
If the BA356 is to house disks with SCSI IDs in the range of 0
through 6, set the BA356 personality module address switches
1 through 7 to off.
•
If the BA356 is to house disks with SCSI IDs in the range of 8
through 14, set BA356 personality module address switches 1
through 3 to on and switches 4 through 7 to off.
If you are using a DWZZB-AA do not replace the personality module
until you attach the cable in the next step.
4. If you are using a DWZZB-AA signal converter, use a BN21K-01
(1-meter) or BN21L-01 (1-meter) cable to connect the single-ended side
of the DWZZB-AA to the BA356 input connector, J A1, on the personality
module. Connector J A1 is on the left side of the personality module as
you face the front of the BA356, and is hidden from normal view. This
connection forms a single-ended bus segment that is terminated by the
DWZZB single-ended termination and the BA356 termination on the
personality module. The use of a 1-meter cable keeps the single-ended
9–16 Configurations Using External Termination or Radial Connections to Non-UltraSCSI Devices
SCSI bus (cable and BA356) under the 3-meter limit to still allow high
speed operation.
If you are using a DWZZB-VW, install it in slot 0 of the BA356.
9.4.1.3 Preparing an UltraSCSI BA356 Storage Shelf for a TruCluster Configuration
An UltraSCSI BA356 storage shelf is connected to a shared UltraSCSI bus,
and provides access to UltraSCSI devices on the internal, single-ended and
wide UltraSCSI bus. The interface between the buses is the DS-BA35X-DA
personality module installed in the UltraSCSI BA356.
To prepare an UltraSCSI BA356 storage shelf for usage on a shared SCSI
bus, follow these steps:
1. Ensure that the BA35X-MJ jumper module is installed behind slot 6
(see Section 9.3.2.1, Figure 9–7, and Figure 9–8).
2. Set the SCSI bus ID switches on the UltraSCSI BA356 personality
module (DS-BA35X-DA, Figure 9–3) as follows:
•
If the UltraSCSI BA356 is to house disks with SCSI IDs in the
range of 0 through 6, set the personality module address switches
S3-1 through S3-7 to OFF.
•
If the UltraSCSI BA356 is to house disks with SCSI IDs in the
range of 8 through 14, set personality module address switches S3-1
through S3-3 to ON and switches S3-4 through S3-7 to OFF.
3. Disable the UltraSCSI BA356 differential termination. Ensure that
personality module (DS-BA35X-DA) switch pack 4 switches S4-1 and
S4-2 are ON (see Figure 9–3).
____________________
Note _____________________
S4-3 and S4-4 are not used on the DS-BA35X-DA.
9.4.2 Connecting Storage Shelves Together
Section 9.4.1 covered the steps necessary to prepare the BA350, BA356, and
UltraSCSI BA356 storage shelves for use on a shared SCSI bus. However,
you will probably need more storage than one storage shelf can provide, so
you will need two storage shelves on the shared SCSI bus.
______________________
Note _______________________
Because the BA350 contains a narrow (8-bit), single-ended SCSI
bus, it only supports SCSI IDs 0 through 7. Therefore, a BA350
Configurations Using External Termination or Radial Connections to
Non-UltraSCSI Devices 9–17
must be used with a BA356 or UltraSCSI BA356 if more than
five disks are required.
The following sections provide the steps needed to connect two storage
shelves and two member systems on a shared SCSI bus:
•
•
•
BA350 and BA356 (Section 9.4.2.1)
Two BA356s (Section 9.4.2.2)
Two UltraSCSI BA356s (Section 9.4.2.3)
9.4.2.1 Connecting a BA350 and a BA356 for Shared SCSI Bus Usage
When you use a BA350 and a BA356 for storage on a shared SCSI bus in a
TruCluster Server configuration, the BA356 must be configured for SCSI
IDs 8 through 14.
To prepare a BA350 and BA356 for shared SCSI bus usage (see Figure 9–9),
follow these steps:
1. Complete the steps in Section 9.4.1.1 and Section 9.4.1.2 to prepare
the BA350 and BA356. Ensure that the BA356 is configured for SCSI
IDs 8 through 14.
2. If either storage shelf will be at the end of the shared SCSI bus, attach
an H879-AA terminator to the H885-AA trilink on the DWZZA or
DWZZB for the storage shelf that will be at the end of the bus. You can
choose either storage shelf to be on the end of the bus.
3. Connect a BN21K or BN21L between the H885-AA trilink on the
DWZZA (BA350) and the H885-AA trilink on the DWZZB (BA356)
4. When the KZPSA-BB or KZPBA-CB host bus adapters have been
installed:
•
If the storage shelves are on the end of the shared SCSI bus, connect
a BN21K (or BN21L) cable between the BN21W-0B Y cables on
the host bus adapters. Connect another BN21K (or BN21L) cable
between the BN21W-0B Y cable with an open connector and the
H8855-AA trilink (on the storage shelf) with an open connector.
•
If the storage shelves are in the middle of the shared SCSI bus,
connect a BN21K (or BN21L) cable between the BN21W-0B Y cable
on each host bus adapter and the H8855-AA trilink on a storage
shelf.
Figure 9–9 shows a two-member TruCluster Server configuration using
a BA350 and a BA356 for storage.
9–18 Configurations Using External Termination or Radial Connections to Non-UltraSCSI Devices
Figure 9–9: BA350 and BA356 Cabled for Shared SCSI Bus Usage
Network
Member
System
1
Member
System
2
Memory
Channel
Interface
Memory Channel
KZPSA-BB (ID 6)
Memory Channel
KZPSA-BB (ID 7)
2
T
2
T
1
1
3
3
BA356
BA350
3
4
DWZZB-VW
DWZZA-VA
4
Clusterwide
/, /usr, /var
ID 9
ID 1
Member 1
Boot Disk
ID 10
ID 11
ID 12
ID 13
ID 2
ID 3
ID 4
ID 5
Member 2
Boot Disk
Data
Disks
Quorum
Disk
Data
disk
ID 14 or
redundant
power
Do not use for
data disk. May
be used for
redundant power
supply.
ID 6
supply
PWR (15)
PWR (7)
ZK-1595U-AI
Table 9–1 shows the components used to create the cluster shown in
Figure 9–9 and Figure 9–10.
Configurations Using External Termination or Radial Connections to
Non-UltraSCSI Devices 9–19
Table 9–1: Hardware Components Used for Configuration Shown in Figure
8–9 and Figure 8–10
Description
Callout Number
1
2
3
BN21W-0B Y cable
H879-AA terminator
BN21K (or BN21L) cablea
H885-AA trilink connector
4
a
The maximum combined length of the BN21K (or BN21L) cables must not exceed 25 meters.
9.4.2.2 Connecting Two BA356s for Shared SCSI Bus Usage
When you use two BA356 storage shelves on a shared SCSI bus in a
TruCluster configuration, one BA356 must be configured for SCSI IDs 0
through 6 and the other configured for SCSI IDs 8 through 14.
To prepare two BA356 storage shelves for shared SCSI bus usage (see
Figure 9–10), follow these steps:
1. Complete the steps of Section 9.4.1.2 for each BA356. Ensure that the
personality module address switches on one BA356 are set to select
SCSI IDs 0 through 6, and that the address switches on the other BA356
personality module are set to select SCSI IDs 8 through 14.
2. If either of the BA356 storage shelves will be on the end of the SCSI bus,
attach an H879-AA terminator to the H885-AA trilink on the DWZZB
for the BA356 that will be on the end of the bus.
3. Connect a BN21K or BN21L cable between the H885-AA trilinks.
4. When the KZPSA-BB or KZPBA-CB host bus adapters have been
installed:
•
•
If the BA356 storage shelves are on the end of the shared SCSI bus,
connect a BN21K (or BN21L) cable between the BN21W-0B Y cables
on the host bus adapters. Connect another BN21K (or BN21L) cable
between the BN21W-0B Y cable with an open connector and the
H8855-AA trilink (on the BA356) with an open connector.
If the BA356s are in the middle of the shared SCSI bus, connect a
BN21K (or BN21L) cable between the BN21W-0B Y cable on each
host bus adapter and the H8855-AA trilink on a BA356 storage shelf.
Figure 9–10 shows a two member TruCluster Server configuration using two
BA356s for storage.
9–20 Configurations Using External Termination or Radial Connections to Non-UltraSCSI Devices
Figure 9–10: Two BA356s Cabled for Shared SCSI Bus Usage
Network
Member
System
1
Member
System
2
Memory
Channel
Interface
Memory Channel
KZPSA-BB (ID 6)
Memory Channel
KZPSA-BB (ID 7)
2
T
2
T
1
1
3
3
BA356
BA356
3
4
DWZZB-VW
DWZZB-VW
Clusterwide
/, /usr, /var
4
ID 9
ID 1
Member 1
Boot Disk
ID 10
ID 11
ID 12
ID 13
ID 2
ID 3
ID 4
ID 5
Member 2
Boot Disk
Data
Disks
Quorum
Disk
Data
disk
ID 14 or
redundant
power
Do not use for
data disk. May
be used for
redundant power
supply.
ID 6
supply
PWR (15)
PWR (7)
ZK-1592U-AI
Table 9–1 shows the components used to create the cluster shown in
Figure 9–10.
9.4.2.3 Connecting Two UltraSCSI BA356s for Shared SCSI Bus Usage
When you use two UltraSCSI BA356 storage shelves on a shared SCSI bus
in a TruCluster configuration, one storage shelf must be configured for SCSI
IDs 0 through 6 and the other configured for SCSI IDs 8 through 14.
Configurations Using External Termination or Radial Connections to
Non-UltraSCSI Devices 9–21
To prepare two UltraSCSI BA356 storage shelves for shared SCSI bus usage,
(see Figure 9–11) follow these steps:
1. Complete the steps of Section 9.4.1.3 for each UltraSCSI BA356. Ensure
that the personality module address switches on one UltraSCSI BA356
are set to select SCSI IDs 0 through 6 and the address switches on
the other UltraSCSI BA356 personality module are set to select SCSI
IDs 8 through 14.
2. You will need two H8861-AA VHDCI trilink connectors. If either of
the UltraSCSI BA356 storage shelves will be on the end of the SCSI
bus, attach an a H8863-AA terminator to one of the H8861-AA trilink
connectors. Install the trilink with the terminator on connector J A1 of
the DS-BA35X-DA personality module of the UltraSCSI BA356 that will
be on the end of the SCSI bus. Install the other H8861-AA trilink on J A1
of the DS-BA35X-DA personality module of the other UltraSCSI BA356.
3. Connect a BN37A VHDCI to VHDCI cable between the H8861-AA
trilink connectors on the UltraSCSI BA356s.
4. When the KZPSA-BBs or KZPBA-CBs are installed:
•
If one of the UltraSCSI BA356s is on the end of the SCSI bus,
install a BN38C (or BN38D) HD68 to VHDCI cable between one of
the BN21W-0B Y cables (on the host bus adapters) and the open
connector on the H8861-AA trilink connector on the DS-BA35X-DA
personality module. Connect the BN21W-0B Y cables on the two
member system host adapters together with a BN21K (or BN21L)
cable.
•
If the UltraSCSI BA356s are in the middle of the SCSI bus, install a
BN38C (or BN38D) HD68 to VHDCI cable between the BN21W-0B
Y cable on each host bus adapter and the open connector on the
H8861-AA trilink connector on the DS-BA35X-DA personality
modules.
9–22 Configurations Using External Termination or Radial Connections to Non-UltraSCSI Devices
Figure 9–11: Two UltraSCSI BA356s Cabled for Shared SCSI Bus Usage
Network
Member
System
1
Member
System
2
Memory
Channel
Interface
Memory Channel
KZPBA-CB (ID 6)
Memory Channel
KZPBA-CB (ID 7)
1
3
T
T
Tru64
UNIX
Disk
2
2
UltraSCSI
BA356
UltraSCSI
BA356
4
5
4
Clusterwide
/, /usr, /var
ID 8
ID 0
ID 1
Member 1
Boot Disk
ID 9
Member 2
Boot Disk
ID 10
ID 11
ID 12
ID 13
ID 2
ID 3
Quorum
Disk
Data
Disks
ID 4
ID 5
ID 4
ID 5
ID 6
Data
disks
Do not use for
data disk. May
be used for
redundant power
supply.
ID 14 or
redundant
power
PWR
PWR
supply
ZK-1598U-AI
Table 9–2 shows the components used to create the cluster shown in
Figure 9–11.
Configurations Using External Termination or Radial Connections to
Non-UltraSCSI Devices 9–23
Table 9–2: Hardware Components Used for Configuration Shown in Figure
9–11
Description
Callout Number
1
2
3
4
BN21W-0B Y cable
H879-AA HD68 terminator
BN38C (or BN38D) cablea
H8861-AA VHDCI trilink connector
BN37A cablea
5
a
The maximum combined length of the BN38C (or BN38D) and BN37A cables on one SCSI bus segment
must not exceed 25 meters.
9.4.3 Cabling a Non-UltraSCSI RAID Array Controller to an Externally
Terminated Shared SCSI Bus
A RAID array controller provides high performance, high availability, and
high connectivity access to SCSI devices through the shared SCSI buses.
Before you connect a RAID controller to a shared SCSI bus, you must install
and configure the disks that the controller will use, and ensure that the
controller has a unique SCSI ID on the shared bus.
You can configure the HSZ20, HSZ40, and HSZ50 RAID array controllers
with one to four SCSI IDs.
Because the HSZ20, HSZ40, and HSZ50 have a wide differential connection
on the host side, you connect them to one of the following differential devices:
•
•
•
KZPSA-BB host bus adapter
KZPBA-CB host bus adapter
Another HSZ20, HSZ40, or HSZ50
______________________
Note _______________________
The HSZ20, HSZ40, and HSZ50 cannot operate at UltraSCSI
speeds when used with the KZPBA-CB.
You can also use a DS-DWZZH-03 or DS-DWZZH-05 UltraSCSI
hub with one of these RAID array controllers and either the
KZPSA-BB or KZPBA-CB host bus adapters. UltraSCSI cables
are required to make the connection to the hub. UltraSCSI speed
is not supported with these RAID array controllers when used
with a hub and the KZPBA-CB host bus adapter.
9–24 Configurations Using External Termination or Radial Connections to Non-UltraSCSI Devices
9.4.3.1 Cabling an HSZ40 or HSZ50 in a Cluster Using External Termination
To connect an HSZ40 or HSZ50 controller to an externally terminated shared
SCSI bus, follow these steps:
1. If the HSZ40 or HSZ50 will be on the end of the shared SCSI bus, attach
an H879-AA terminator to an H885-AA trilink connector.
2. Attach an H885-AA trilink connector to each RAID controller port.
Attach the H885-AA trilink connector with the terminator to the
controller that will be on the end of the shared SCSI bus.
3. If you are using dual-redundant RAID array controllers, install a
BN21-K or BN21L cable (a BN21L-0B is a 0.15-meter cable) between
the H885-AA trilink connectors on the controllers.
4. When the host bus adapters (KZPSA-BB or KZPBA-CB) have been
installed, connect the host bus adapters and RAID array controllers
together with BN21K or BN21L cables as follows:
•
Both member systems are on the ends of the bus: Attach a BN21K or
BN21L cable from the BN21W-0B Y cable on each host bus adapter
to the RAID array controller(s).
•
RAID array controller is on the end of the bus: Connect a BN21K
(or BN21L) cable from the BN21W-0B Y cable on one host bus
adapter to the BN21W-0B Y cable on the other host bus adapter.
Attach another BN21K (or BN21L) cable from the open BN21W-0B
Y cable connector to the open H885-AA connector on the RAID array
controller.
Figure 9–12 shows two AlphaServer systems in a TruCluster Server
configuration with dual-redundant HSZ50 RAID controllers in the middle
of the shared SCSI bus. Note that the SCSI bus adapters are KZPSA-BB
PCI-to-SCSI adapters. They could be KZPBA-CB host bus adapters without
changing any cables.
Configurations Using External Termination or Radial Connections to
Non-UltraSCSI Devices 9–25
Figure 9–12: Externally Terminated Shared SCSI Bus with Mid-Bus HSZ50
RAID Array Controllers
Network
Member
System
1
Member
System
2
Memory
Channel
Interface
Memory Channel
KZPSA-BB (ID 6)
Memory Channel
KZPSA-BB (ID 7)
T
T
1
1
3
4
4
3
3
2
2
HSZ50
Controller A
HSZ50
Controller B
ZK-1596U-AI
Table 9–3 shows the components used to create the cluster shown in
Figure 9–12 and Figure 9–13.
Figure 9–13 shows two AlphaServer systems in a TruCluster Server
configuration with dual-redundant HSZ50 RAID controllers at the end of
the shared SCSI bus. As with Figure 9–12, the SCSI bus adapters are
KZPSA-BB PCI-to-SCSI adapters. They could be KZPBA-CB host bus
adapters without changing any cables.
9–26 Configurations Using External Termination or Radial Connections to Non-UltraSCSI Devices
Figure 9–13: Externally Terminated Shared SCSI Bus with HSZ50 RAID
Array Controllers at Bus End
Network
Member
System
1
Member
System
2
Memory
Channel
Interface
Memory Channel
KZPSA-BB (ID 6)
Memory Channel
KZPSA-BB (ID 7)
3
3
T
1
1
4
4
3
2
2
T
HSZ50
Controller A
HSZ50
Controller B
ZK-1597U-AI
Table 9–3 shows the components used to create the cluster shown in
Figure 9–12 and Figure 9–13.
Table 9–3: Hardware Components Used for Configuration Shown in Figure
8–12 and Figure 8–13
Description
Callout Number
1
2
3
BN21W-0B Y cable
H879-AA terminator
BN21K (or BN21L) cablea
H885-AA trilink connector
b
4
a
b
The maximum combined length of the BN21K (or BN21L) cables must not exceed 25 meters.
The cable between the H885-AA trilink connectors on the HSZ50s could be a BN21L-0B, a 0.15-meter cable.
9.4.3.2 Cabling an HSZ20 in a Cluster using External Termination
To connect a SWXRA-Z1 (HSZ20 controller) to a shared SCSI bus, follow
these steps:
1. Referring to the RAID Array 310 Deskside Subsystem (SWXRA-ZX)
Hardware User ’s Guide, open the SWXRA-Z1 cabinet, locate the SCSI
bus converter board, and:
•
Remove the five differential terminator resistor SIPs.
Configurations Using External Termination or Radial Connections to
Non-UltraSCSI Devices 9–27
•
Ensure that the W1 and W2 jumpers are installed to enable the
single-ended termination on one end of the bus.
___________________ Note
___________________
The RAID Array 310 SCSI bus converter board is the
same logic board used in the DWZZB signal converter.
2. Attach an H885-AA trilink connector to the SCSI input connector (on
the back of the cabinet).
3. Use a BN21K or BN21L cable to connect the trilink connector to a
trilink connector or BN21W-0B Y cable attached to a differential SCSI
controller, another storage shelf, or the differential end of a signal
converter.
4. Terminate the differential bus by attaching an H879-AA terminator to
the H885-AA trilink connector or BN21W-0B Y cable at each end of
the shared SCSI bus.
Ensure that all devices that make up the shared SCSI bus are connected,
and that there is a terminator at each end of the shared SCSI bus.
9.4.4 Cabling an HSZ40 or HSZ50 RAID Array Controller in a Radial
Configuration with an UltraSCSI Hub
You may have an HSZ40 or HSZ50 that you wish to keep when you upgrade
to a newer AlphaServer system. You can connect an HSZ40 or HSZ50 to an
UltraSCSI hub in a radial configuration, but even if the host bus adapter is a
KZPBA-CB, it will not operate at UltraSCSI speed with the HSZ40 or HSZ50.
To configure a dual-redundant HSZ40 or HSZ50 RAID array controller and
an UltraSCSI hub in a radial configuration, follow these steps:
1. You will need two H885-AA trilink connectors. Install an H879-AA
terminator on one of the trilinks.
2. Attach the trilink with the terminator to the controller that you want
to be on the end of the shared SCSI bus. Attach an H885-AA trilink
connector to the other controller.
3. Install a BN21K or BN21L cable between the H885-AA trilink
connectors on the two controllers. The BN21L-0B is a 0.15-meter cable.
4. If you are using a DS-DWZZH-05:
•
Verify that the fair arbitration switch is in the Fair position to
enable fair arbitration (see Section 3.6.1.2.2)
•
Ensure that the W1 jumper is removed to select wide addressing
mode (see Section 3.6.1.2.3)
9–28 Configurations Using External Termination or Radial Connections to Non-UltraSCSI Devices
5. Install the UltraSCSI hub in:
•
A StorageWorks UltraSCSI BA356 shelf (which has the required
180-watt power supply).
•
A non-UltraSCSI BA356 which has been upgraded to the 180-watt
power supply with the DS-BA35X-HH option.
6. If you are using a:
•
DS-DWZZH-03: Install a BN38C (or BN38D) HD to VHDCI cable
between any DS-DWZZH-03 port and the open connector on the
H885-AA trilink connector (on the RAID array controller).
•
DS-DWZZH-05: Install a BN38C (or BN38D) cable between the
DWZZH-05 controller port and the open trilink connector on HSZ40
or HSZ50 controller.
___________________ Note
___________________
Ensure that the HSZ40 or HSZ50 SCSI IDs match the
DS-DWZZH-05 controller port IDs (SCSI IDs 0-6)
7. When the host bus adapters (KZPSA-BB or KZPBA-CB) have been
installed in the member systems, for a:
•
DS-DWZZH-03: Install a BN38C (or BN38D) HD68 to VHDCI cable
between the KZPBA-CB or KZPSA-BB host bus adapter to each of
the other two DS-DWZZH-03 ports.
•
DS-DWZZH-05: Install a BN38C (or BN38D) HD68 to VHDCI
cable between the KZPBA-CB or KZPSA-BB host bus adapter on
each system to a port on the DWZZH hub. Ensure that the host bus
adapter SCSI ID matches the SCSI ID assigned to the DWZZH-05
port it is cabled to (12, 13, 14, and 15).
Figure 9–14 shows a sample configuration with radial connection of
KZPSA-BB PCI-to-SCSI adapters, DS-DWZZH-03 UltraSCSI hub, and an
HSZ50 RAID array controller. Note that the KZPSA-BBs could be replaced
with KZPBA-CB UltraSCSI adapters without any changes in cables.
Configurations Using External Termination or Radial Connections to
Non-UltraSCSI Devices 9–29
Figure 9–14: TruCluster Server Cluster Using DS-DWZZH-03, SCSI Adapter
with Terminators Installed, and HSZ50
1
KZPSA-BB
DS-DWZZH-03
T
T
1
AlphaServer
Member
T
T
System 1
2
4
2
HSZ50
HSZ50
1
KZPSA-BB
T
T
AlphaServer
Member
System 2
3
ZK-1415U-AI
Table 9–4 shows the components used to create the cluster shown in
Figure 9–14.
Table 9–4: Hardware Components Used in Configuration Shown in Figure
9–14
Description
Callout Number
b
BN38C cablea
1
2
3
H885-AA HD68 trilink connector
H879-AA HD68 terminator
BN21K or BN21L cableb
4
a
b
The maximum length of the BN38C cable on one SCSI bus segment must not exceed 25 meters.
The maximum combined length of the BN38C and BN21K (or BN21L) cables on the storage SCSI bus
segment must not exceed 25 meters.
Figure 9–15 shows a sample configuration that uses KZPSA-BB SCSI
adapters, a DS-DWZZH-05 UltraSCSI hub, and an HSZ50 RAID array
controller.
9–30 Configurations Using External Termination or Radial Connections to Non-UltraSCSI Devices
Figure 9–15: TruCluster Server Cluster Using KZPSA-BB SCSI Adapters, a
DS-DWZZH-05 UltraSCSI Hub, and an HSZ50 RAID Array Controller
AlphaServer
Member
System 1
T
KZPSA-BB
1
1
KZPSA-BB
DS-DWZZH-05
T
T
1
T
AlphaServer
Member
T
T T
System 2
2
4
2
HSZ50
HSZ50
1
1
KZPSA-BB
KZPSA-BB
T
T
T
AlphaServer
Member
System 3
AlphaServer
Member
System 4
3
ZK-1449U-AI
______________________
Note _______________________
The systems shown in Figure 9–15 use KZPSA-BB SCSI adapters.
They could be KZPBA-CB UltraSCSI adapters without changing
any cables in the configuration.
Table 9–4 shows the components used to create the cluster shown in
Figure 9–15.
Configurations Using External Termination or Radial Connections to
Non-UltraSCSI Devices 9–31
10
Configuring Systems for External
Termination or Radial Connections to
Non-UltraSCSI Devices
This chapter describes how to prepare the systems for a TruCluster Server
cluster when there is a need for external termination or radial connection to
non-UltraSCSI RAID array controllers (HSZ40 and HSZ50). This chapter
does not provide detailed information about installing devices; it describes
only how to set up the hardware in the context of the TruCluster Server
product. Therefore, you must have the documentation that describes how
to install the individual pieces of hardware. This documentation should
arrive with the hardware.
All systems in the cluster must be connected via the Memory Channel
cluster interconnect. Not all members must be connected to a shared SCSI
bus. We recommend placing the clusterwide root (/), /usr, and /var file
systems, all member boot disks, and the quorum disk (if provided) on shared
SCSI buses. All configurations covered in this manual assume the use of a
shared SCSI bus.
Before proceeding further, review Section 4.1, Section 4.2, and the first two
paragraphs of Section 4.3.
10.1 TruCluster Server Hardware Installation Using PCI
SCSI Adapters
The following sections describe how to install the KZPSA-BB or KZPBA-CB
host bus adapters and configure them into TruCluster Server clusters using
both methods of termination — the preferred method of radial connection
with internal termination used with the HSZ40 and HSZ50 RAID array
controllers, and the old method of external termination.
It is assumed that you have already configured and cabled your storage
subsystems as described in Chapter 9. When the system hardware
(KZPSA-BB or KZPBA-CB host bus adapters, Memory Channel adapters,
hubs (if necessary), cables, and network adapters) have been installed,
you can connect your host bus adapter to the UltraSCSI hub or storage
subsystem.
Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices 10–1
Follow the steps in Table 10–1 to start the TruCluster Server hardware
installation procedure. You can save time by installing the Memory Channel
adapters, redundant network adapters (if applicable), and KZPSA-BB or
KZPBA-CB SCSI adapters all at the same time.
Follow the directions in the referenced documentation, or the steps in the
referenced tables for the particular SCSI host bus adapter, returning to the
appropriate table when you have completed the steps in the referenced table.
_____________________ Caution
_____________________
Static electricity can damage modules and electronic components.
We recommend using a grounded antistatic wrist strap and a
grounded work surface when handling modules.
Table 10–1: Configuring TruCluster Server Hardware for Use with a PCI
SCSI Adapter
Action
Refer to:
Step
Chapter 5a
1
Install the Memory Channel module(s),
cables, and hub(s) (if a hub is required).
2
3
Install Ethernet or FDDI network
adapters.
User ’s guide for the applicable
Ethernet or FDDI adapter,
and the user ’s guide for the
applicable system
Install ATM adapters if using ATM.
Chapter 7 and ATMworks 350
Adapter Installation and Service
Install a KZPSA-BB PCI SCSI adapter or
KZPBA-CB UltraSCSI adapter for each
shared SCSI bus in each member system.
Internally terminated host bus adapter
for radial connection to DWZZH
UltraSCSI hub:
Section 10.1.1 and Table 10–2
Externally terminated host bus adapter: Section 10.1.2 Table 10–3
a
If you install additional KZPSA-BB or KZPBA-CB SCSI adapters or an extra network adapter at this time,
delay testing the Memory Channel until you have installed all hardware.
10.1.1 Radial Installation of a KZPSA-BB or KZPBA-CB Using Internal
Termination
Use this method of cabling member systems and shared storage in a
TruCluster Server cluster if you are using a DWZZH UltraSCSI hub. You
must reserve at least one hub port for shared storage.
10–2 Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices
The DWZZH-series UltraSCSI hubs are designed to allow more separation
between member systems and shared storage. Using the UltraSCSI hub also
improves the reliability of the detection of cable faults.
A side benefit is the ability to connect the member systems’ SCSI adapter
directly to a hub port without external termination. This simplifies the
configuration by reducing the number of cable connections.
A DWZZH UltraSCSI hub can be installed in:
•
A StorageWorks UltraSCSI BA356 shelf (which has the required
180-watt power supply).
•
A non-UltraSCSI BA356 that has been upgraded to the 180-watt power
supply with the DS-BA35X-HH option.
An UltraSCSI hub only receives power and mechanical support from the
storage shelf. There is no SCSI bus continuity between the DWZZH and
storage shelf.
The DWZZH contains a differential to single-ended signal converter for each
hub port (sometimes referred to as a DWZZA on a chip, or DOC chip). The
single-ended sides are connected together to form an internal single-ended
SCSI bus segment. Each differential SCSI bus port is terminated internal to
the DWZZH with terminators that cannot be disabled or removed.
Power for the DWZZH termination (termpwr) is supplied by the host bus
adapter or RAID array controller connected to the DWZZH port. If the
member system or RAID array controller is powered down, or the cable is
removed from the host bus adapter, RAID array controller, or hub port, the
loss of termpwr disables the hub port without affecting the remaining hub
ports or SCSI bus segments. This is similar to removing a Y cable when
using external termination.
The other end of the SCSI bus segment is terminated by the KZPSA-BB
or KZPBA-CB onboard termination resistor SIPs, or a trilink
connector/terminator combination installed on the HSZ40 or HSZ50.
The KZPSA-BB PCI-to-SCSI bus adapter:
•
•
•
Is installed in a PCI slot of the supported member system (see
Section 2.4.1).
Is a fast, wide differential adapter with only a single port, so only one
differential shared SCSI bus can be connected to a KZPSA-BB adapter.
Operates at fast or slow speed and is compatible with narrow or wide
SCSI. The fast speed is 10 MB/sec for a narrow SCSI bus and 20 MB/sec
for a wide SCSI bus. The KZPSA-BB must be set to fast speed for
TruCluster Server.
Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices 10–3
_____________________ Note _____________________
You may have problems if the member system supports the
bus_probe_algorithm console variable and it is not set to
new. See Section 2.4.1.
The KZPBA-CB UltraSCSI host adapter:
•
Is a high-performance PCI option connecting the PCI-based host system
to the devices on a 16-bit, ultrawide differential SCSI bus.
•
•
Is a single-channel, ultrawide differential adapter.
Operates at the following speeds:
–
–
–
–
5 MB/sec narrow SCSI at slow speed
10 MB/sec narrow SCSI at fast speed
20 MB/sec wide differential SCSI
40 MB/sec wide differential UltraSCSI
______________________
Note _______________________
Even though the KZPBA-CB is an UltraSCSI device, it has an
HD68 connector.
Use the steps in Table 10–2 to set up a KZPSA-BB or KZPBA-CB host bus
adapter for a TruCluster Server cluster that uses radial connection to a
DWZZH UltraSCSI hub with an HSZ40 or HSZ50 RAID array controller.
Table 10–2: Installing the KZPSA-BB or KZPBA-CB for Radial Connection
to a DWZZH UltraSCSI Hub
Action
Refer to:
Step
1
Ensure that the KZPSA-BB internal termination
resistors, Z1, Z2, Z3, Z4, and Z5 are installed.
Section 10.1.4.4,
Figure 10–1, and KZPSA
PCI-to-SCSI Storage
Adapter Installation
and User ’s Guide
Ensure that the eight KZPBA-CB internal
Section 4.3.3.3,
termination resistor SIPs, RM1-RM8 are installed. Figure 4–1, and
KZPBA-CB PCI-to-Ultra
SCSI Differential Host
Adapter User ’s Guide
10–4 Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices
Table 10–2: Installing the KZPSA-BB or KZPBA-CB for Radial Connection
to a DWZZH UltraSCSI Hub (cont.)
Action
Refer to:
Step
2
Power down the system. Install a KZPSA-BB
KZPSA PCI-to-SCSI
PCI-to-SCSI adapter or KZPBA-CB UltraSCSI host Storage Adapter
adapter in the PCI slot corresponding to the logical
bus to be used for the shared SCSI bus. Ensure
that the number of adapters are within limits for
Installation and User ’s
Guide and KZPBA-CB
PCI-to-Ultra SCSI
the system, and that the placement is acceptable. Differential Host Adapter
User ’s Guide
3
Install a BN38C cable between the KZPBA-BB
or KZPBA-CB UltraSCSI host adapter
and a DWZZH port.
_____________________
Notes
_____________________
The maximum length of a SCSI bus segment is 25 meters, including the
bus length internal to the adapter and storage devices.
One end of the BN38C cable is 68-pin high density. The other end is
68-pin VHDCI. The DWZZH accepts the 68-pin VHDCI connector.
The number of member systems in the cluster has to be one less than
the number of DWZZH ports.
4
Power up the system, and update the system
SRM console firmware and KZPSA-BB host
bus adapter firmware from the latest Alpha
Systems Firmware Update CD-ROM.
Firmware release
notes for the system
(Section 4.2) and
Section 10.1.4.5
______________________ Note ______________________
The SRM console firmware includes the ISP1020/1040-based PCI
option firmware, which includes the KZPBA-CB. When you update the
SRM console firmware, you are enabling the KZPBA-CB firmware to
be updated. On a powerup reset, the SRM console loads KZPBA-CB
adapter firmware from the console system flash ROM into NVRAM for
all Qlogic ISP1020/1040-based PCI options, including the KZPBA-CB
PCI-to-Ultra SCSI adapter.
5
Use the show config and show device console Section 10.1.3 and
commands to display the installed devices
and information about the KZPSA-BBs or
KZPBA-CBs on the AlphaServer systems. Look
for KZPSA or pk* in the display to determine
which devices are KZPSA-BBs. Look for QLogic
ISP1020 in the show config display and
isp in the show device display to determine
which devices are KZPBA-CBs.
Example 10–1 through
Example 10–4
Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices 10–5
Table 10–2: Installing the KZPSA-BB or KZPBA-CB for Radial Connection
to a DWZZH UltraSCSI Hub (cont.)
Action
Refer to:
Step
6
Use the show pk* or show isp* console
commands to determine the status of the
KZPSA-BB or KZPBA-CB console environment
variables, and then use the set console
command to set the KZPSA-BB bus speed to
fast, termination power to on, and the KZPSA
or KZPBA-CB SCSI bus ID.
Section 10.1.4.1 through
Section 10.1.4.3 and
Example 10–6 through
Example 10–9
_____________________
Notes
_____________________
Ensure that the SCSI ID that you use is distinct from all other SCSI IDs
on the same shared SCSI bus. If you do not remember the other SCSI
IDs, or do not have them recorded, you must determine these SCSI IDs.
If you are using a DS-DWZZH-05, you cannot use SCSI ID 7 for a
KZPSA-BB or KZPBA-CB host bus adapter; SCSI ID 7 is reserved for
DS-DWZZH-05 use.
If you are using a DS-DWZZH-05 and fair arbitration is enabled, you
must use the SCSI ID assigned to the hub port the adapter is to be
connected to.
You will have problems if you have two or more SCSI adapters at the
same SCSI ID on any one SCSI bus.
7
8
Repeat steps 1 through 6 for any other KZPSA-BBs
or KZPBA-CBs to be installed on this shared
SCSI bus on other member systems.
Connect a DS-DWZZH-03 or DS-DWZZH-05
to an HSZ40 or HSZ50
Section 9.4.4
10.1.2 Installing a KZPSA-BB or KZPBA-CB Using External
Termination
Use the steps in Table 10–3 to set up a KZPSA-BB or KZPBA-CB for a
TruCluster Server cluster using the old method of external termination
and Y cables.
10–6 Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices
Table 10–3: Installing a KZPSA-BB or KZPBA-CB for use with External
Termination
Action
Refer to:
Step
1
Remove the KZPSA-BB internal termination
resistors, Z1, Z2, Z3, Z4, and Z5.
Section 10.1.4.4,
Figure 10–1, and KZPSA
PCI-to-SCSI Storage
Adapter Installation
and User ’s Guide
Remove the eight KZPBA-CB internal termination Section 4.3.3.3,
resistor SIPs, RM1-RM8.
Figure 4–1, and
KZPBA-CB PCI-to-Ultra
SCSI Differential Host
Adapter User ’s Guide
2
Power down the member system. Install
a KZPSA-BB PCI-to-SCSI bus adapter or
KZPBA-CB UltraSCSI host adapter in the PCI
KZPSA PCI-to-SCSI
Storage Adapter
Installation and User ’s
slot corresponding to the logical bus to be used for Guide and KZPBA-CB
the shared SCSI bus. Ensure that the number
of adapters are within limits for the system,
and that the placement is acceptable.
PCI-to-Ultra SCSI
Differential Host Adapter
User ’s Guide
3
4
Install a BN21W-0B Y cable on each KZPSA-BB
or KZPBA-CB host adapter.
Install an H879-AA terminator on one leg of the
BN21W-0B Y cable of the member system that
will be on the end of the shared SCSI bus.
5
Power up the system, and update the system
SRM console firmware and KZPSA-BB host
bus adapter firmware from the latest Alpha
Systems Firmware Update CD-ROM.
Firmware release
notes for the system
(Section 4.2) and
Section 10.1.4.5
______________________ Note ______________________
The SRM console firmware includes the ISP1020/1040-based PCI
option firmware, which includes the KZPBA-CB. When you update the
SRM console firmware, you are enabling the KZPBA-CB firmware to
be updated. On a powerup reset, the SRM console loads KZPBA-CB
adapter firmware from the console system flash ROM into NVRAM for
all Qlogic ISP1020/1040-based PCI options, including the KZPBA-CB
PCI-to-Ultra SCSI adapter.
6
Use the show config and show device console Section 10.1.3 and
commands to display the installed devices
and information about the KZPSA-BBs or
KZPBA-CBs on the AlphaServer systems. Look
for KZPSA or pk* in the display to determine
which devices are KZPSA-BBs. Look for QLogic
ISP1020 in the show config display and
isp in the show device display to determine
which devices are KZPBA-CBs.
Example 10–1 through
Example 10–4
Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices 10–7
Table 10–3: Installing a KZPSA-BB or KZPBA-CB for use with External
Termination (cont.)
Action
Refer to:
Step
7
Use the show pk* or show isp* console
commands to determine the status of the
KZPSA-BB or KZPBA-CB console environment
variables, and then use the set console
command to set the KZPSA-BB bus speed to
fast, termination power to on, and the KZPSA
or KZPBA-CB SCSI bus ID.
Section 10.1.4.1 through
Section 10.1.4.3 and
Example 10–6 through
Example 10–9
_____________________
Notes
_____________________
Ensure that the SCSI ID that you use is distinct from all other SCSI IDs
on the same shared SCSI bus. If you do not remember the other SCSI
IDs, or do not have them recorded, you must determine these SCSI IDs.
You will have problems if you have two or more SCSI adapters at the
same SCSI ID on any one SCSI bus.
8
9
Repeat steps 1 through 7 for any other KZPSA-BBs
or KZPBA-CBs to be installed on this shared
SCSI bus on other member systems.
Install the remaining SCSI bus hardware needed Section 9.4
for storage (DWZZA(B), RAID array controllers,
storage shelves, cables, and terminators).
BA350 storage shelf.
Section 9.3.1,
Section 9.4.1.1, and
Section 9.4.2.1
Non-UltraSCSI BA356 storage shelf.
Ultra BA356 storage shelf.
Section 9.3.2.1,
Section 9.4.1.2, and
Section 9.4.2.2
Section 9.3.2.2,
Section 9.4.1.3, and
Section 9.4.2.3
HSZ40 or HSZ50 RAID array controller
Section 9.4.3
Chapter 8
10
Install the tape device hardware and cables
on the shared SCSI bus as follows:
TZ88
Section 8.1
Section 8.2
Section 8.3
Section 8.4
Section 8.5
Section 8.6
TZ89
Compaq 20/40 GB DLT Tape Drive
TZ885
TZ887
TL891/TL892 MiniLibrary
10–8 Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices
Table 10–3: Installing a KZPSA-BB or KZPBA-CB for use with External
Termination (cont.)
Action
Refer to:
Section 8.7
Step
TL890 with TL891/TL892
TL894
Section 8.8
TL895
Section 8.9
TL893/TL896
Section 8.10
TL881/TL891 DLT MiniLibraries
Compaq ESL9326D Enterprise Library
Section 8.11
Section 8.12
_____________________
Notes
_____________________
If you install tape devices on the shared SCSI buses, ensure that you
understand how the particular tape device(s) affect the shared SCSI bus.
The TL893, TL894, TL895, TL896, and ESL9326D have long internal
SCSI cables; therefore, they cannot be externally terminated with a
trilink/terminator combination.
These tape libraries must be on the end of the shared SCSI bus.
We recommend that tape devices be placed on a separate shared SCSI
bus.
10.1.3 Displaying KZPSA-BB and KZPBA-CB Adapters with the show
Console Commands
Use the show config and show device console commands to display
system configuration. Use the output to determine which devices are
KZPSA-BBs or KZPBA-CBs, and to determine their SCSI bus IDs.
Example 10–1 shows the output from the show config console command
on an AlphaServer 4100 system.
Example 10–1: Displaying Configuration on an AlphaServer 4100
P00>>> show config
Compaq Computer Corporation
AlphaServer 4x00
Console V5.1-3 OpenVMS PALcode V1.19-14, Tru64 UNIX PALcode V1.21-22
Module
Type
Rev
Name
System Motherboard
0
0
0
0
0
0000
0000
0000
0000
0000
mthrbrd0
mem0
mem1
mem2
mem3
Memory
Memory
Memory
Memory
64 MB SYNC
64 MB SYNC
64 MB SYNC
64 MB SYNC
Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices 10–9
Example 10–1: Displaying Configuration on an AlphaServer 4100 (cont.)
CPU (4MB Cache)
CPU (4MB Cache)
Bridge (IOD0/IOD1)
PCI Motherboard
3
3
600
8
0000
0000
0021
0000
cpu0
cpu1
iod0/iod1
saddle0
Bus 0 iod0 (PCI0)
Slot
Option Name
PCEB
S3 Trio64/Trio32
DECchip 21040-AA
DEC KZPSA
Type
4828086 0005
88115333 0000
21011
81011
181011
Rev
Name
pceb0
vga0
tulip0
pks1
mc0
1
2
3
4
5
0024
0000
000B
DEC PCI MC
Bus 1 pceb0 (EISA Bridge connected to iod0, slot 1)
Slot
Option Name
Type
Rev
Name
Bus 0 iod1 (PCI1)
Slot
Option Name
NCR 53C810
NCR 53C810
QLogic ISP1020
QLogic ISP1020
DEC KZPSA
Type
11000
11000
10201077 0005
10201077 0005
81011
Rev
0002
0002
Name
ncr0
ncr1
isp0
isp1
pks0
1
2
3
4
5
0000
Example 10–2 shows the output from the show device console command
entered on an AlphaServer 4100 system.
Example 10–2: Displaying Devices on an AlphaServer 4100
P00>>> show
device
polling ncr0 (NCR 53C810) slot 1, bus0 PCI, hose 1 SCSI Bus ID 7
dka500.5.0.1.1
Dka500
RRD45
1645
polling ncr1 (NCR 53C810) slot 2, bus0 PCI, hose 1 SCSI Bus ID 7
dkb0.0.0.2.1
dkb100.1.0.2.1
DKb0
DKb100
RZ29B
RZ29B
0007
0007
polling isp0 (QLogic ISP1020) slot 3, bus 0 PCI, hose 1 SCSI Bus ID
7
dkc0.0.0.3.1
dkc1.0.0.3.1
dkc2.0.0.3.1
dkc3.0.0.3.1
dkc4.4.0.3.1
dkc5.0.0.3.1
dkc6.0.0.3.1
dkc100.1.0.3.1
dkc200.2.0.3.1
dkc300.3.0.3.1
DKc0
DKc1
DKc2
DKc3
DKc4
DKc5
DKc6
DKc100
DKc200
DKc300
HSZ70
HSZ70
HSZ70
HSZ70
HSZ70
HSZ70
HSZ70
RZ28M
RZ28M
RZ28
V70Z
V70Z
V70Z
V70Z
V70Z
V70Z
V70Z
0568
0568
442D
polling isp1 (QLogic ISP1020) slot 4, bus 0 PCI, hose 1 SCSI Bus ID
7
dkd0.0.0.4.1
dkd1.0.0.4.1
dkd2.0.0.4.1
DKd0
DKd1
DKd2
HSZ50-AX X29Z
HSZ50-AX X29Z
HSZ50-AX X29Z
10–10 Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices
Example 10–2: Displaying Devices on an AlphaServer 4100 (cont.)
dkd100.1.0.4.1
dkd200.1.0.4.1
dkd300.1.0.4.1
DKd100
DKd200
DKd300
RZ26N
RZ26
RZ26N
0568
392A
0568
polling kzpsa0 (DEC KZPSA) slot 5, bus 0 PCI, hose 1 TPwr 1 Fast 1 Bus ID 7
kzpsa0.7.0.5.1
dke100.1.0.5.1
dke200.2.0.5.1
dke300.3.0.5.1
dke TPwr 1 Fast 1 Bus ID 7 L01 A11
DKe100
DKe200
DKe300
RZ28
RZ26
RZ26L
442D
392A
442D
polling floppy0 (FLOPPY) pceb IBUS hose 0
dva0.0.0.1000.0 DVA0 RX23
polling kzpsa1 (DEC KZPSA) slot 4, bus 0 PCI, hose 0 TPwr 1 Fast 1 Bus ID 7
kzpsa1.7.0.4.1
dkf100.1.0.5.1
dkf200.2.0.5.1
dkf300.3.0.5.1
polling tulip0
ewa0.0.0.3.0
dkf TPwr 1 Fast 1 Bus ID 7 E01 A11
DKf100
DKf200
DKf300
RZ26
RZ28
RZ26
392A
442D
392A
(DECchip 21040-AA) slot 3, bus 0 PCI, hose 0
00-00-F8-21-0B-56 Twisted-Pair
Example 10–3 shows the output from the show config console command
entered on an AlphaServer 8200 system.
Example 10–3: Displaying Configuration on an AlphaServer 8200
>>> show config
Name
Type
Rev
Mnemonic
TLSB
4++
5+
KN7CC-AB
MS7CC
KFTIA
8014
5000
2020
0000
0000
0000
kn7cc-ab0
ms7cc0
kftia0
8+
C0 Internal PCI connected to kftia0
pci0
isp0
isp1
0+ QLogic ISP1020 10201077
1+ QLogic ISP1020 10201077
2+ DECchip 21040-AA 21011
4+ QLogic ISP1020 10201077
5+ QLogic ISP1020 10201077
6+ DECchip 21040-AA 21011
0001
0001
0023 tulip0
0001
0001
0023 tulip1
isp2
isp3
C1 PCI connected to kftia0
0+ KZPAA
11000
0001 kzpaa0
1+ QLogic ISP1020 10201077
0005
isp4
2+ KZPSA
3+ KZPSA
4+ KZPSA
7+ DEC PCI MC
81011
81011
81011
181011
0000 kzpsa0
0000 kzpsa1
0000 kzpsa2
000B
mc0
Example 10–4 shows the output from the show device console command
entered on an AlphaServer 8200 system.
Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices 10–11
Example 10–4: Displaying Devices on an AlphaServer 8200
>>> show device
polling for units on isp0, slot0, bus0, hose0...
polling for units on isp1, slot1, bus0, hose0...
polling for units on isp2, slot4, bus0, hose0...
polling for units on isp3, slot5, bus0, hose0...
polling for units kzpaa0, slot0, bus0, hose1...
pke0.7.0.0.1
dke0.0.0.0.1
dke200.2.0.0.1
dke400.4.0.0.1
kzpaa4
DKE0
DKE200
DKE400
SCSI Bus ID 7
RZ28
442D
442D
0064
RZ28
RRD43
polling for units isp4, slot1, bus0, hose1...
dkf0.0.0.1.1
dkf1.0.0.1.1
dkf2.0.0.1.1
dkf3.0.0.1.1
dkf4.0.0.1.1
dkf5.0.0.1.1
dkf6.0.0.1.1
dkf100.1.0.1.1
dkf200.2.0.1.1
dkf300.3.0.1.1
DKF0
DKF1
DKF2
DKF3
DKF4
DKF5
DKF6
DKF100
DKF200
DKF300
HSZ70
HSZ70
HSZ70
HSZ70
HSZ70
HSZ70
HSZ70
RZ28M
RZ28M
RZ28
V70Z
V70Z
V70Z
V70Z
V70Z
V70Z
V70Z
0568
0568
442D
polling for units on kzpsa0, slot 2, bus 0, hose1...
kzpsa0.4.0.2.1
dkg0.0.0.2.1
dkg1.0.0.2.1
dkg2.0.0.2.1
dkg100.1.0.2.1
dkg200.2.0.2.1
dkg300.3.0.2.1
dkg
DKG0
DKG1
DKG2
DKG100
DKG200
DKG300
TPwr 1 Fast 1 Bus ID 7
HSZ50-AX X29Z
L01 A11
HSZ50-AX X29Z
HSZ50-AX X29Z
RZ26N
RZ28
RZ26N
0568
392A
0568
polling for units on kzpsa1, slot 3, bus 0, hose1...
kzpsa1.4.0.3.1
dkh100.1.0.3.1
dkh200.2.0.3.1
dkh300.3.0.3.1
dkh
TPwr 1 Fast 1 Bus ID 7
L01 A11
DKH100
DKH200
DKH300
RZ28
RZ26
442D
392A
442D
RZ26L
polling for units on kzpsa2, slot 4, bus 0, hose1...
kzpsa2.4.0.4.1
dki100.1.0.3.1
dki200.2.0.3.1
dki300.3.0.3.1
dki
TPwr 1 Fast 1 Bus ID 7
L01 A10
DKI100
DKI200
DKI300
RZ26
RZ28
RZ26
392A
442C
392A
10–12 Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices
10.1.4 Displaying Console Environment Variables and Setting the
KZPSA-BB and KZPBA-CB SCSI ID
The following sections show how to use the show console command to display
the pk* and isp* console environment variables and set the KZPSA-BB and
KZPBA-CB SCSI ID on various AlphaServer systems. Use these examples
as guides for your system.
Note that the console environment variables used for the SCSI options vary
from system to system. Also, a class of environment variables (for example,
pk* or isp*) may show both internal and external options.
Compare the following examples with the devices shown in the show
config and show dev examples to determine which devices are KZPSA-BBs
or KZPBA-CBs on the shared SCSI bus.
10.1.4.1 Displaying KZPSA-BB and KZPBA-CB pk* or isp* Console Environment
Variables
To determine the console environment variables to use, execute the show
pk* and show isp* console commands.
Example 10–5 shows the pk console environment variables for an
AlphaServer 4100.
Example 10–5: Displaying the pk* Console Environment Variables on an
AlphaServer 4100 System
P00>>>show pk*
pka0_disconnect
pka0_fast
pka0_host_id
1
1
7
pkb0_disconnect
pkb0_fast
pkb0_host_id
1
1
7
pkc0_host_id
7
pkc0_soft_term
diff
pkd0_host_id
7
pkd0_soft_term
on
pke0_fast
pke0_host_id
1
7
Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices 10–13
Example 10–5: Displaying the pk* Console Environment Variables on an
AlphaServer 4100 System (cont.)
pke0_termpwr
1
pkf0_fast
pkf0_host_id
pkf0_termpwr
1
7
1
Compare the show pk* command display in Example 10–5 with the
show config command in Example 10–1 and the show dev command
in Example 10–2. Note that there are no pk* devices in either display.
Example 10–2 shows:
•
•
•
The NCR 53C810 SCSI controllers as ncr0 and ncr1 with disk DKa and
DKb (pka and pkb)
The Qlogic ISP1020 devices (KZPBA-CBs) as isp0 and isp1 with disks
DKc and DKd (pkc and pkd)
The KZPSA-BBs with disks DKe and DKf (pke and pkf)
Example 10–5 shows two pk*0_soft_term environment variables;
pkc0_soft_term which is on, and pkd0_soft_term which is diff.
The pk*0_soft_term environment variable applies to systems using the
QLogic ISP1020 SCSI controller, which implements the 16-bit wide SCSI
bus and uses dynamic termination.
The QLogic ISP1020 module has two terminators, one for the 8 low bits and
one for the high 8 bits. There are five possible values for pk*0_soft_term:
•
•
•
•
•
off — Turns off both the low 8 bits and high 8 bits
low — Turns on the low 8 bits and turns off the high 8 bits
high — Turns on the high 8 bits and turns off the low 8 bits
on — Turns on both the low 8 bits and high 8 bits
diff — Places the bus in differential mode
The KZPBA-CB is a Qlogic ISP1040 module, and its termination is
determined by the presence or absence of internal termination resistor SIPs
RM1-RM8. Therefore, the pkb0_soft_term environment variable has no
meaning and it may be ignored.
Example 10–6 shows the use of the show isp console command to display
the console environment variables for KZPBA-CBs on an AlphaServer 8x00.
10–14 Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices
Example 10–6: Displaying Console Variables for a KZPBA-CB on an
AlphaServer 8x00 System
P00>>> show isp*
isp0_host_id
7
isp0_soft_term
on
isp1_host_id
7
isp1_soft_term
on
isp2_host_id
7
isp2_soft_term
on
isp3_host_id
7
isp3_soft_term
on
isp5_host_id
7
isp5_soft_term
diff
Both Example 10–3 and Example 10–4 show five isp devices; isp0, isp1,
isp2, isp3, and isp4. In Example 10–6, the show isp* console command
shows isp0, isp1, isp2, isp3, and isp5.
The console code that assigns console environment variables counts every I/O
adapter including the KZPAA, which is the device after isp3, and therefore
logically isp4 in the numbering scheme. The show isp console command
skips over isp4 because the KZPAA is not a QLogic 1020/1040 class module.
Example 10–3 and Example 10–4 show that isp0, isp1, isp2, and isp3
are on the internal KFTIA PCI bus and not on a shared SCSI bus. Only
isp5, the KZPBA-CB, is on a shared SCSI bus. The other three shared
SCSI buses use KZPSA-BBs.
Example 10–7 shows the use of the show pk console command to display
the console environment variables for KZPSA-BBs on an AlphaServer 8x00.
Example 10–7: Displaying Console Variables for a KZPSA-BB on an
AlphaServer 8x00 System
P00>>> show pk*
pka0_fast
1
pka0_host_id
pka0_termpwr
7
on
pkb0_fast
1
pkb0_host_id
pkb0_termpwr
7
on
Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices 10–15
Example 10–7: Displaying Console Variables for a KZPSA-BB on an
AlphaServer 8x00 System (cont.)
pkc0_fast
1
pkc0_host_id
pkc0_termpwr
7
on
10.1.4.2 Setting the KZPBA-CB SCSI ID
After you determine the console environment variables for the KZPBA-CBs
on the shared SCSI bus, use the set console command to set the SCSI ID.
For a TruCluster Server cluster, you will most likely have to set the SCSI
ID for all KZPBA-CB UltraSCSI adapters except one. If you are using a
DS-DWZZH-05 with fair arbitration enabled, you will have to set the SCSI
IDs for all KZPBA-CB UltraSCSI adapters.
______________________
Note _______________________
You will have problems if you have two or more SCSI adapters at
the same SCSI ID on any one SCSI bus.
If you are using a DS-DWZZH-05, you cannot use SCSI ID 7
for a KZPBA-CB UltraSCSI adapter; SCSI ID 7 is reserved for
DS-DWZZH-05 use.
If DS-DWZZH-05 fair arbitration is enabled, the SCSI ID of the
host adapter must match the SCSI ID assigned to the hub port.
Mismatching or duplicating SCSI IDs will cause the hub to hang.
Use the set console command as shown in Example 10–8 to set the
KZPBA-CB SCSI ID. In this example, the SCSI ID is set for KZPBA-CB pkc
on the AlphaServer 4100 shown in Example 10–5.
Example 10–8: Setting the KZPBA-CB SCSI Bus ID
P00>>> show pkc0_host_id
7
P00>>> set pkc0_host_id 6
P00>>> show pkc0_host_id
6
10–16 Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices
10.1.4.3 Setting KZPSA-BB SCSI Bus ID, Bus Speed, and Termination Power
If the KZPSA-BB SCSI ID is not correct, or if it was reset to 7 by the
firmware update utility, or you need to change the KZPSA-BB speed, or
enable termination power, use the set console command.
______________________
Note _______________________
All KZPSA-BB host bus adapters should be enabled to generate
termination power.
Set the SCSI bus ID with the set command as shown in the following
example:
>>> set pkn_0_host_id #
The n specifies which KZPSA-BB the environment variables apply to. You
obtain the n value from the show device and show pk* console commands.
The number sign (#) is the SCSI bus ID for the KZPSA.
Set the bus speed with the set command as shown in the following example:
>>> set pkn0_fast #
The number sign (#) specifies the bus speed. Use a 0 for slow and a 1 for fast.
Enable SCSI bus termination power with the set command as shown in
the following example:
>>> set pkn0_termpwr on
Example 10–9 shows how to determine the present SCSI ID, bus speed,
and the status of termination power, and then set the KZPSA-BB SCSI ID
to 6 and bus speed to fast for pkb0.
Example 10–9: Setting KZPSA-BB SCSI Bus ID and Speed
P00>>> show pkb*
pkb0_fast 0
pkb0_host_id 7
pkb0_termpwr on
P00>>> set pkb0_host_id 6
P00>>> set pkb0_fast 1
P00>>> show pkb0_host_id
6
P00>>> show pkb0_fast
1
Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices 10–17
10.1.4.4 KZPSA-BB and KZPBA-CB Termination Resistors
The KZPSA-BB internal termination is disabled by removing termination
resistors Z1 through Z5, as shown in Figure 10–1.
Figure 10–1: KZPSA-BB Termination Resistors
Z1 − Z5 Termination
Resistor SIPs
The KZPBA-CB internal termination is disabled by removing the
termination resistors RM1-RM8 as shown in Figure 4–1.
10.1.4.5 Updating the KZPSA-BB Adapter Firmware
You must check, and update as necessary, the system and host bus adapter
firmware. The firmware may be out of date. Read the firmware release
notes from the AlphaSystems Firmware Update CD-ROM for the applicable
system/SCSI adapter.
If the Standard Reference Manual (SRM) console or KZPSA-BB firmware
is not current, boot the Loadable Firmware Update (LFU) utility from the
Alpha Systems Firmware Update CD-ROM. Choose the update entry from
the list of LFU commands. LFU can update all devices or any particular
device you select.
When you boot the Systems Firmware Update CD-ROM, you can
read the firmware release notes. After booting has completed, enter
read_rel_notes at the UPD> prompt. You can also copy and print the
release notes as shown in Section 4.2.
To update the firmware, boot the LFU utility from the Alpha Systems
Firmware Update CD-ROM.
It is not necessary to use the -flag option to the boot command. Insert
the Alpha Systems Firmware Update CD-ROM and boot. For example, to
boot from dka600:
P00>>> boot dka600
10–18 Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices
The boot sequence provides firmware update overview information. Use
Return to scroll the text, or press Ctrl/C to skip the text.
After the overview information has been displayed, the name of the default
boot file is provided. If it is the correct boot file, press Return at the
Bootfile: prompt. Otherwise, enter the name of the file you wish to boot
from.
The firmware images are copied from the CD-ROM and the LFU help
message shown in the following example is displayed:
*****Loadable Firmware Update Utility*****
-------------------------------------------------------------
Function
Description
-------------------------------------------------------------
Display
Exit
Displays the system’s configuration table.
Done exit LFU (reset).
List
Lists the device, revision, firmware name and
update revision
Readme
Update
Lists important release information.
Replaces current firmware with loadable data
image.
Verify
? or Help
Compares loadable and hardware images.
Scrolls this function table.
The list command indicates, in the device column, which devices it can
update.
Use the update command to update all firmware, or you can designate a
specific device to update; for example, KZPSA-BB pkb0:
UPD> update pkb0
After updating the firmware and verifying this with the verify command,
reset the system by cycling the power.
Configuring Systems for External Termination or Radial Connections
to Non-UltraSCSI Devices 10–19
A
Worldwide ID to Disk Name Conversion
Table
Table A–1: Converting Storageset Unit Numbers to Disk Names
WWID
User
Device Name
dskn
File System HSG80
or Disk
Unit
Define
Iden-
tifier
(UDID)
Tru64 UNIX
disk
Cluster root
(/)
/usr
/var
Member 1
boot disk
Member 2
boot disk
Member 3
boot disk
Member 4
boot disk
Quorum disk
Worldwide ID to Disk Name Conversion Table A–1
Index
A
BA370
DS-DWZZH-03 installed in, 2–9,
3–9
bootdef_dev, 6–46, 6–48, 6–50,
6–51, 6–54
setting, 6–46, 6–48, 6–49, 6–51,
6–54
bus hung message, 2–8
bus_probe_algorithm, 2–6
buses
ACS V8.5, 2–4
arbitrated loop, 6–7
ATL
TL893, 8–40, 8–41
TL896, 8–40, 8–41
ATM
atmconfig command, 7–4
connecting cables, 7–4
installation, 7–1
LANE, 7–1
data paths, 3–5
extending differential, 9–2
narrow data path, 3–5
speed, 3–5
terminating, 3–7, 9–5, 9–8
wide data path, 3–5
atmconfig command, 7–4
availability
increasing, 4–3
B
C
BA350, 9–9
preparing, 9–15
cables
BC12N-10, 5–7
BN39B-04, 5–7
BN39B-10, 5–7
ESL9326D, 8–68
supported, 2–10
preparing for shared SCSI usage,
9–15
termination, 9–3, 9–15
BA356, 9–9
DS-DWZZH-03 installed in, 2–9,
3–9, 4–7, 9–29, 10–3
DS-DWZZH-05 installed in, 3–10
jumper, 9–9, 9–12
personality module address
switches, 9–11
preparing, 9–15, 9–17
preparing for shared SCSI usage,
9–16
SCSI ID selection, 9–16
selecting SCSI IDs, 9–11
termination, 9–3, 9–9, 9–12
cabling
Compaq 20/40 GB DLT Tape Drive,
8–10
DS-TZ89N-TA, 8–8
DS-TZ89N-VW, 8–7
ESL9326D, 8–65, 8–68
TL881/891 DLT MiniLibrary, 8–54,
8–58
TL890, 8–24
TL891, 8–20, 8–24
Index–1
TL892, 8–20, 8–24
TL893, 8–47
cabling, 8–10
capacity, 8–9
TL894, 8–34
cartridges, 8–9
TL895, 8–40
connectors, 8–9
TL896, 8–47
TZ885, 8–13
setting SCSI ID, 8–9
configuration
restrictions, 2–5
CONFIGURATION RESTORE
command, 6–33
TZ887, 8–16
TZ88N-TA, 8–4
TZ88N-VA, 8–3
configuring base unit as slave, 8–26,
8–61
changing HSG80 failover modes, 6–55
cluster
connections to HSG80, 6–55
connectors
supported, 2–11
console variable
bus_probe_algorithm, 2–6
expanding, 3–7, 9–6
increasing availability, 4–3
planning, 4–2
cluster interconnects
increasing availability, 4–2
clusterwide file systems
allocating a disk for, 1–4
command
D
data path
atmconfig, 7–4
for buses, 3–5
CONFIGURATION RESTORE,
6–33
emxmgr, 6–59
emxmgr -d, 6–59
emxmgr -m, 6–59
default SCSI IDs
ESL9326D, 8–66
TL881/TL891, 8–53
TL890, 8–29
TL891, 8–29
emxmgr -t, 6–60
TL892, 8–29
TL893, 8–43
TL894, 8–30
TL895, 8–37
init, 6–24, 6–45, 6–48, 6–50
SAVE_CONFIGURATION, 6–33
set bootdef_dev, 6–48, 6–50
SET FAILOVER COPY =
THIS_CONTROLLER, 1–14
SET MULTIBUS_FAILOVER
COPY = THIS_CONTROLLER,
3–18
show config, 4–9t, 10–5t, 10–7t
show device, 4–9t, 10–5t, 10–7t
SHOW THIS_CONTROLLER,
6–31
TL896, 8–44
device name, 6–40
device unit number, 6–40
setting, 6–41
diagnostics
Memory Channel, 5–12
differential SCSI buses
description of, 3–4
differential transmission
definition, 3–4
wwidmgr, 6–40
wwidmgr -clear, 6–41
wwidmgr -quickset, 6–42
wwidmgr -show, 6–24, 6–42, 6–45
Compaq 20/40 GB DLT Tape Drive,
8–9
disk devices
restrictions, 2–7
setting up, 3–16, 9–14
disk placement
Index–2
clusterwide /usr, 1–10
clusterwide /var, 1–10
clusterwide root, 1–10
member boot, 1–10
quorum, 1–10
configurations, 3–15
description, 2–9
fair arbitration, 3–10
installed in, 3–10, 3–11
internal termination, 3–9
radial disconnect, 2–9
SBB, 3–10
SCSI ID, 3–10
termpwr, 3–9
transfer rate, 2–9
DS-TZ89N-TA
cabling, 8–8
disklabel, 6–53
displaying device information
KZPBA-CB, 4–9t, 10–5t, 10–7t
KZPSA-BB, 10–5t, 10–7t
DLT
Compaq 20/40 GB DLT Tape Drive,
8–9
TZ885, 8–13
TZ887, 8–15
setting SCSI ID, 8–8
DS-TZ89N-VW
DLT MiniLibrary
cabling, 8–7
Configuring TL881/TL891 as slave,
8–61
Configuring TL891 as slave, 8–26
TL881, 8–48
setting SCSI ID, 8–5
dual-redundant controllers, 1–14
DWZZA
incorrect hardware revision, 2–8
termination, 9–3, 9–16
upgrade, 2–8
TL891, 8–48
DS-BA356
DS-DWZZH-03 installed in, 2–9,
3–9, 4–7, 9–29, 10–3
DS-DWZZH-05 installed in, 3–10
DS-BA35X-DA personality module,
3–3, 3–5, 4–8, 9–2, 9–3
DS-DWZZH-03, 3–9
bus connectors, 3–9
bus isolation, 2–9
DWZZB
termination, 9–3, 9–16
DWZZH-03
( See DS-DWZZH-03 )
E
emxmgr, 6–59
displaying adapters, 6–59
displaying target ID mapping, 6–59
displaying topology, 6–60
use, 6–59, 6–61
using interactively, 6–62
enterprise library
description, 2–9
installed in, 2–9, 3–9, 4–7, 9–29,
10–3
internal termination, 3–9
radial disconnect, 2–9
SBB, 3–9
( See ESL9326D )
environment variable
bootdef_dev, 6–46, 6–48, 6–49,
6–51, 6–54
N, 6–41
wwid, 6–41
SCSI ID, 3–9
support on, 3–9
termpwr, 3–9
transfer rate, 2–9
DS-DWZZH-05, 3–9
bus connectors, 3–10
bus isolation, 2–9
ESA12000
Index–3
configuring, 2–5
port configuration, 2–5
replacing controllers of, 6–32
transparent failover mode, 2–5
unit configuration, 2–5
ESL9000 series tape library
( See ESL9326D )
ESL9326D
supported configurations, 6–8
switch installation, 6–15
terminology, 6–4
topology, 6–5, 6–61
file
/var/adm/messages, 6–26
firmware
ESL9326D, 8–66
cables, 8–68
cabling, 8–65, 8–68
capacity, 8–65
obtaining release notes, 4–4
reset system for update, 10–19
update CD-ROM, 4–4
updating, 10–18
firmware, 8–66
internal cabling, 8–67
number of drives, 8–65
part numbers, 8–65
SCSI connectors, 8–68
setting SCSI IDs, 8–66
tape cartridges, 8–65
tape drives, 8–65
updating KZPSA, 10–18
FL_Port, 6–5
G
GBIC, 6–16
termination, 8–68
upgrading, 8–65
H
hardware components
Fibre Channel, 2–3
SCSI adapters, 2–6
F
SCSI cables, 2–10
F_Port, 6–5
SCSI signal converters, 2–8
storage shelves, 9–9
fabric, 6–5
failover mode
changing, 6–55
multiple-bus, 6–55
set nofailover, 6–56
transparent, 6–55
Fibre Channel
arbitrated loop, 6–7
data rates, 6–4
distance, 6–4
F_Port, 6–5
terminators, 2–11
trilink connectors, 2–11
hardware configuration
bus termination, 3–7, 9–5
disk devices, 3–16, 9–14
hardware requirements for, 2–1
hardware restrictions for, 2–1
requirements, 3–1, 9–1
SCSI bus adapters, 2–6
SCSI bus speed, 3–5
fabric, 6–6
FL_Port, 6–5
SCSI cables, 2–10
frame, 6–4
N_Port, 6–5
NL_Port, 6–5
point-to-point, 6–6
restrictions, 2–3
SCSI signal converters, 9–2
storage shelves, 3–16, 9–14
supported cables, 2–1
supported terminators, 2–1
supported trilinks, 2–1
Index–4
supported Y cables, 2–1
terminators, 2–11
( See also hardware
configuration )
trilink connectors, 2–11
host bus adapters
ATM adapter, 7–1
KGPSA, 6–22
( See KGPSA, KZPBA-CB,
KZPSA-BB )
HSG80 controller
KZPSA, 10–3
MC2, 5–10
MC2 cables, 5–9
Memory Channel, 5–5
Memory Channel cables, 5–7
Memory Channel hub, 5–6
optical converter, 5–6
optical converter cables, 5–10
switch, 6–15
ACS, 2–4
changing failover modes, 6–55
configuring, 2–5, 6–26
multiple-bus failover, 6–28
obtaining the worldwide name of,
6–31
port configuration, 2–5
port_n_topology, 6–29
replacing, 6–32
internal cabling
ESL9326D, 8–67
TL893, 8–44
resetting offsets, 6–55
setting controller values, 6–26,
6–28
TL896, 8–44
J
transparent failover mode, 2–5
unit configuration, 2–5
HSZ failover
multiple-bus, 1–15
transparent, 1–14
HSZ20 controller
jumpers
MC1 and MC1.5 (CCMAA), 5–2
MC2 (CCMAB), 5–3
MC2 (CCMLB), 5–5
Memory Channel, 5–2
and shared SCSI bus, 9–24
HSZ40 controller, 1–14
and shared SCSI bus, 9–24
HSZ50 controller, 1–14
and shared SCSI bus, 9–24
HSZ70 controller, 1–14
and fast wide differential SCSI, 3–3
HSZ80 controller, 1–14
hwmgr, 6–52
K
KGPSA
GLM, 6–23
installing, 6–22
mounting bracket, 6–22
obtaining the worldwide name of,
6–26
KZPBA-CB
displaying device information,
4–9t, 10–5t, 10–7t
restrictions, 2–6
termination resistors, 4–9t, 10–4t,
10–7t
I
I/O buses
number of, 2–6
initialize, 6–24, 6–45, 6–48, 6–50
installation, 3–16
use in cluster, 4–6, 10–2
Index–5
KZPSA-BB
diagnostics, 5–12
installation, 5–2, 5–5
interconnect, 2–3
jumpers, 5–2
displaying device information,
10–5t, 10–7t
installation, 10–3
restrictions, 2–6
versions, 2–2
Memory Channel diagnostics
mc_cable, 5–12
mc_diag, 5–12
Memory Channel hub
installation, 5–6
Memory Channel interconnects
restrictions, 2–2
setting bus speed, 10–17
setting SCSI ID, 10–17
setting termination power, 10–17
termination resistors, 10–4t, 10–7t
updating firmware, 10–18
use in cluster, 10–2
setting up, 5–1
message
L
bus hung, 2–8
MiniLibrary
TL881, 8–48
TL891, 8–48
minimum cluster configuration, 1–5
MUC, 8–42
setting SCSI ID, 8–43
MUC switch functions
TL893, 8–42
TL896, 8–42
multi-unit controller
( See MUC )
multimode fibre, 6–16, 7–1
multiple-bus failover, 1–15, 3–18,
3–22, 6–28
changing from transparent failover,
6–55
LAN emulation
( See LANE )
LANE, 7–1
LFU, 10–18
booting, 10–18
starting, 10–18
updating firmware, 10–18
link cable
installation, 5–7
Loadable Firmware Update utility
( See LFU )
Logical Storage Manager
( See LSM )
LSM
mirroring across SCSI buses, 1–11
mirroring clusterwide /usr, 1–12
mirroring clusterwide /var, 1–12
mirroring clusterwide data disks,
1–12
NSPOF, 3–18
setting, 6–28, 6–56
N
M
N_Port, 6–5
NL_Port, 6–5
mc_cable, 5–12
mc_diag, 5–12
node name, 6–31
non-Ultra BA356 storage shelf
preparing, 9–15
member systems
improving performance, 4–2
increasing availability, 4–2
requirements, 2–1
Memory Channel
NSPOF, 1–13, 3–18
Index–6
replacing controllers of, 6–32
transparent failover mode, 2–5
unit configuration, 2–5
radial connection
bus termination, 3–8
UltraSCSI hub, 3–8
RAID, 1–13
O
optical cable, 6–16
optical converter
cable connection, 5–6
installation, 5–6
RAID array controllers
advantages of use, 3–17
and shared SCSI bus, 9–24
preparing, 9–24
using in ASE, 9–24
replacing
HSG80 controller, 6–32
requirements
SCSI bus, 3–1, 9–1
reset, 6–24, 6–45
resetting offsets, 6–55
restrictions, 2–5
disk devices, 2–7
KZPBA-CB adapters, 2–6
KZPSA adapters, 2–6
Memory Channel interconnects,
2–2
P
part numbers
ESL9326D, 8–65
partitioned storagesets, 3–18
performance
improving, 4–2
personality module, 3–3, 9–13
( See also signal converters )
planning the hardware configuration,
4–2
point-to-point, 6–6
port name, 6–31
powering up
TL881/891 DLT MiniLibrary, 8–62
preparing storage shelves
BA350, 9–15
BA350 and BA356, 9–18
BA356, 9–16, 9–20
UltraSCSI BA356, 9–17, 9–21
Prestoserve
SCSI bus adapters, 2–6
S
cannot be used in a cluster, 4–3
SAVE_CONFIGURATION
command, 6–33
SC connector
( See subscriber connector )
SCSI
number of devices supported, 3–2
SCSI bus with BA350 and BA356,
9–18
Q
quorum disk
and LSM, 1–5
configuring, 1–5
number of votes, 1–5
SCSI bus with Two BA356s, 9–20
SCSI bus with two UltraSCSI BA356s,
9–21
SCSI buses
( See shared SCSI buses )
SCSI cables
R
RA8000
configuring, 2–5
port configuration, 2–5
Index–7
( See cables )
TZ887, 8–15
requirement, 2–10
SCSI controllers
TZ88N-TA, 8–4
TZ88N-VA, 8–2
bus speed for, 3–5
SCSI ID selection, 9–17
BA356, 9–16
setting SCSI IDs
ESL9326D, 8–66
setting the SCSI ID
TL881/891 DLT MiniLibrary, 8–53
shared SCSI buses, 4–3
adding devices, 9–6
assigning SCSI IDs, 3–6
cable length restrictions, 3–6
connecting devices, 3–7, 9–6
device addresses, 3–5
differential, 3–4
SCSI IDs
BA350, 9–9
BA350 storage shelves, 9–15
BA356, 9–11, 9–16
HSZ20 controller, 9–24
HSZ40 controller, 9–24
HSZ50 controller, 9–24
in BA356, 9–11
in UltraSCSI BA356, 9–13
RAID subsystem controllers, 9–24
requirement, 3–5
UltraSCSI BA356, 9–13, 9–17
SCSI targets
number of, 2–6, 4–3
requirements, 3–2
single-ended, 3–4
using trilink connectors, 9–6
using Y cables, 9–6
shared storage
number of, 2–5
SCSI terminators
supported, 2–11
SCSI-2 bus, 3–5
BA350 storage shelf, 9–15
increasing capacity, 4–2, 4–3
non-UltraSCSI BA356 storage
shelf, 9–15
SCSI-3, 6–4
selecting BA356 disk SCSI IDs, 9–11
selecting UltraSCSI BA356 disk SCSI
IDs, 9–13
set bootdef_dev, 6–48, 6–50, 6–54
setting bus speed
RAID subsystem array controller,
9–24
UltraSCSI BA356 storage shelf,
9–15, 9–17
shortwave, 6–23
SHOW THIS_CONTROLLER
KZPSA, 10–17
setting SCSI ID
command, 6–31
Compaq 20/40 GB DLT Tape Drive,
8–9
DS-TZ89N-TA, 8–8
DS-TZ89N-VW, 8–5
KZPSA, 10–17
MUC, 8–43
TL881/891 DLT MiniLibrary, 8–63
TL891, 8–18
TL892, 8–18
TL893, 8–43
TL894, 8–30
TL896, 8–43
signal converters, 9–2
creating differential bus, 9–2
differential I/O module, 9–2
differential termination, 9–3
DS-BA35X-DA personality module,
3–5, 9–4
extending differential bus length,
9–2
fast SCSI bus speed, 9–2
overview, 9–2
requirement, 9–2
restrictions, 2–8
TZ885, 8–13
Index–8
SBB, 9–2
single-ended termination, 9–3
standalone, 9–2
terminating, 9–2
termination, 9–3
terminating the shared bus, 3–7,
9–5
UltraSCSI BA356, 9–14
termination resistors
KZPBA-CB, 4–9t, 10–4t, 10–7t
KZPSA, 10–4t, 10–7t
KZPSA-BB, 10–7t
single-ended SCSI buses
description of, 3–4
single-ended transmission
definition, 3–4
terminators
supported, 2–11
storage shelves, 9–8, 9–9, 9–13
attaching to shared SCSI bus, 9–8,
9–13
TL881, 8–48
TL881/891 DLT MiniLibrary
cabling, 8–54, 8–58
capacity, 8–49, 8–51
components, 8–49
BA350, 9–9
BA356, 9–9
configuring base unit as slave, 8–61
models, 8–49
performance, 8–51
overview, 9–8, 9–13
setting up, 3–16, 9–14
subscriber connector, 6–16
switch
powering up, 8–62
10Base-T Ethernet connection,
6–15
setting the SCSI ID, 8–53, 8–63
TL890
cabling, 8–24
default SCSI IDs, 8–29
powering up, 8–28
setting SCSI ID, 8–29
TL891, 8–18, 8–48
changing password, 6–21
changing user names, 6–21
front panel, 6–15, 6–18
GBIC, 6–16
installing, 6–15
cabling, 8–20, 8–24
configuring as slave, 8–26
default SCSI IDs, 8–20, 8–29
setting SCSI ID, 8–18, 8–19, 8–29
shared SCSI usage, 8–18
TL892, 8–18
interface module, 6–16
overview, 6–15
setting Ethernet IP address, 6–18
setting switch name, 6–21
telnet session, 6–21
system reset, 6–24, 6–45
cabling, 8–20, 8–24
configuring as slave, 8–26
default SCSI IDs, 8–20, 8–29
setting SCSI ID, 8–18, 8–19, 8–29
shared SCSI usage, 8–18
TL893, 8–40, 8–41
T
table of connections, 6–55
termination, 9–13
BA356, 9–11
cabling, 8–44, 8–47
MUC switch functions, 8–42
setting SCSI ID, 8–43
TL894
DWZZA, 9–16
DWZZB, 9–16
ESL9326D, 8–68
cabling, 8–34
Index–9
setting SCSI ID, 8–30
TL895
preparing for shared SCSI usage,
9–17
cabling, 8–40
SCSI ID selection, 9–17
selecting SCSI IDs, 9–13
termination, 9–14
UltraSCSI host adapter
host input connector, 3–3
with non-UltraSCSI BA356, 3–3
with UltraSCSI BA356, 3–3
UltraSCSI hubs, 3–8
unshielded twisted pair
( See UTP )
TL896, 8–40, 8–41
cabling, 8–44, 8–47
MUC switch functions, 8–42
setting SCSI ID, 8–43
transparent failover, 1–14, 3–17
changing to multiple-bus failover,
6–55
trilink connectors
connecting devices with, 9–6
requirement, 2–11
supported, 2–11
upgrade
DWZZA, 2–8
upgrading
TZ88, 8–1
ESL9326D, 8–65
utility
hwmgr, 6–52
wwidmgr, 6–45, 6–49
UTP, 7–1
versions, 8–1
TZ885, 8–13
cabling, 8–13
setting SCSI ID, 8–13
TZ887, 8–15
cabling, 8–16
setting SCSI ID, 8–15
TZ88N-TA, 8–1
cabling, 8–4
setting SCSI ID, 8–4
TZ88N-VA, 8–1
cabling, 8–3
setting SCSI ID, 8–2
TZ89, 8–5
V
/var/adm/messages, 6–26
variable
( See environment variable )
Very High Density Cable Interconnect
( See VHDCI )
VHDCI, 3–3
acronym defined, 3–3
HSZ70 host connector, 3–3
U
UltraSCSI BA356
W
disable termination, 9–17
DS-BA35X-DA personality module,
3–3
fast narrow SCSI drives, 3–3
fast wide SCSI drives, 3–3
jumper, 9–14
personality module address
switches, 9–13
power supply, 3–3
WorldWide ID Manager
( See wwidmgr )
worldwide name
description, 6–25
wwidmgr, 6–40
-clear, 6–41
-quickset, 6–42
-show, 6–24, 6–42, 6–45
preparing, 9–15, 9–17
Index–10
connecting devices with, 9–6
supported, 2–10
Y
Y cables
Index–11
How to Order Tru64 UNIX Documentation
You can order documentation for the Tru64 UNIX operating system and related
products at the following Web site:
http://www.businesslink.digital.com/
If you need help deciding which documentation best meets your needs, see the
Tru64 UNIX Documentation Overview or call 800-344-4825 in the United States
and Canada. In Puerto Rico, call 787-781-0505. In other countries, contact your
local Compaq subsidiary.
If you have access to Compaq’s intranet, you can place an order at the following
Web site:
http://asmorder.nqo.dec.com/
The following table provides the order numbers for the Tru64 UNIX operating system
documentation kits. For additional information about ordering this and related
documentation, see the Documentation Overview or contact Compaq.
Name
Order Number
Tru64 UNIX Documentation CD-ROM
Tru64 UNIX Documentation Kit
End User Documentation Kit
QA-6ADAA-G8
QA-6ADAA-GZ
QA-6ADAB-GZ
QA-6ADAC-GZ
QA-6ADAD-GZ
QA-6ADAE-GZ
QA-6ADAG-GZ
QA-6ADAF-GZ
Startup Documentation Kit
General User Documentation Kit
System and Network Management Documentation Kit
Developer ’s Documentation Kit
Reference Pages Documentation Kit
Reader’s Comments
TruCluster Server
Hardware Configuration
AA-RHGWB-TE
Compaq welcomes your comments and suggestions on this manual. Your input will help us to write
documentation that meets your needs. Please send your suggestions using one of the following methods:
•
•
•
This postage-paid form
Internet electronic mail: [email protected]
Fax: (603) 884-0120, Attn: UBPG Publications, ZKO3-3/Y32
If you are not using this form, please be sure you include the name of the document, the page number, and
the product name and version.
P lea se r a te th is m a n u a l:
Excellent
Good
Fair
Poor
Accuracy (software works as manual says)
Clarity (easy to understand)
Organization (structure of subject matter)
Figures (useful)
Examples (useful)
Index (ability to find topic)
Usability (ability to access information quickly)
P lea se list er r or s you h a ve fou n d in th is m a n u a l:
Page
Description
_________
_________
_________
_________
_______________________________________________________________________________________
_______________________________________________________________________________________
_______________________________________________________________________________________
_______________________________________________________________________________________
Ad d ition a l com m en ts or su ggestion s to im p r ove th is m a n u a l:
___________________________________________________________________________________________________
___________________________________________________________________________________________________
___________________________________________________________________________________________________
___________________________________________________________________________________________________
Wh a t ver sion of th e softw a r e d escr ibed by th is m a n u a l a r e you u sin g?
_______________________
Name, title, department ___________________________________________________________________________
Mailing address __________________________________________________________________________________
Electronic mail ___________________________________________________________________________________
Telephone ________________________________________________________________________________________
Date _____________________________________________________________________________________________
Do Not Cut or Tear - Fold Here and Tape
NO POSTAGE
NECESSARY IF
MAILED IN THE
UNITED STATES
FIRST CLASS MAIL PERMIT NO. 33 MAYNARD MA
POSTAGE WILL BE PAID BY ADDRESSEE
COMPAQ COMPUTER CORPORATION
UBPG PUBLICATIONS MANAGER
ZKO3-3/Y32
110 SPIT BROOK RD
NASHUA NH 03062-2698
Do Not Cut or Tear - Fold Here
|