Compaq StorageWorks TM RAID Array 3000 Pedestal Storage Subsystem Manuel de service

TruCluster ServerCluster Hardware ConfigurationPart Number: AA-RHGWD-TEJune 2001Product Version: TruCluster Server Version 5.1AOperating System and Ve

8.12.2Preparing a TL881 or TL891 MiniLibrary for Shared SCSIBus Use ...8–618.12.2.1Preparing

______________________ Note _______________________If you are using Fibre Channel storage, see Chapter 6.Before you connect devices to a shared SCSI b

• Cluster interconnectsYou need only one cluster interconnect in a cluster. For TruClusterServer Version 5.1A, the cluster interconnect can be the Mem

Table 4–1: Planning Your ConfigurationTo increase: You can:Application performance Increase the number of member systems.I/O performance Increase the

5. Mount the CD-ROM as follows (/dev/disk/cdrom0c is used as anexample CD-ROM drive):# mount -rt cdfs -o noversion /dev/disk/cdrom0cc /mnt6. Copy the

TruCluster Server clusters using the preferred method of radial connectionwith internal termination.______________________ Note ______________________

Table 4–2: Configuring TruCluster Server Hardware (cont.)StepAction Refer to:4Update the system SRM consolefirmware from the latest Alpha SystemsFirmw

The DWZZH contains a differential to single-ended signal converter foreach hub port (which is sometimes referred to as a DWZZA on a chip,or DOC chip).

Make sure that your storage shelves or RAID array subsystems are set upbefore completing this portion of an installation.Use the steps in Table 4–3 to

Table 4–3: Installing the KZPBA-CB for Radial Connection to a DWZZHUltraSCSI Hub (cont.)StepAction Refer to:_____________________ Notes ______________

Example 4–1: Displaying Configuration on an AlphaServer DS20 (cont.)CPU 1 Alpha 21264-4 500 MHz SROM Revision: V1.82Bcache size: 4 MBCore LogicCchip D

9.1.4.2Setting the KZPBA-CB SCSI ID ... ... ... 9–169.1.4.3Setting KZPSA-BB SCSI Bus ID, Bus Speed, andTermination Power . .. ... ...

Example 4–1: Displaying Configuration on an AlphaServer DS20 (cont.)Bus 02 Slot 01: NCR 53C875pkb0.7.0.2001.0 SCSI Bus ID 7Bus 02 Slot 02: DE500-AA Ne

Example 4–3 shows the output from the show config console commandentered on an AlphaServer 8200 system.Example 4–3: Displaying Configuration on an Alp

Example 4–4: Displaying Devices on an AlphaServer 8200 (cont.)dkf4.0.0.1.1 DKF4 HSZ70 V70Zdkf5.0.0.1.1 DKF5 HSZ70 V70Zdkf6.0.0.1.1 DKF6 HSZ70 V70Zdkf1

4.3.3.1 Displaying KZPBA-CB pk* or isp* Console Environment VariablesTo determine the console environment variables to use, execute the showpk* and sh

• on — Turns on both low 8 bits and high 8 bits• diff — Places the bus in differential modeThe KZPBA-CB is a Qlogic ISP1040 module, and its terminatio

4.3.3.2 Setting the KZPBA-CB SCSI IDAfter you determine the console environment variables for the KZPBA-CBson the shared SCSI bus, use the set console

Figure 4–1: KZPBA-CB Termination ResistorsInternal Wide DeviceConnector J2Internal Narrow DeviceConnector P2SCSI Bus TerminationResistors RM1-RM8ZK-14

5Setting Up the Memory Channel ClusterInterconnectThis chapter describes Memory Channel configuration restrictions, anddescribes how to set up the Mem

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-o

Table 5–1: MC1 and MC1.5 J4 Jumper ConfigurationIf hub mode is: Jumper: Example:Standard J4 Pins 1 to 2123Virtual: VH0 J4 Pins 2 to 3123Virtual: VH1 N

10.4.4Cabling a Non-UltraSCSI RAID Array Controller to anExternally Terminated Shared SCSI Bus . ... 10–2510.4.4.1Cabling an HSZ40 or

space. The configuration change is propagated to the other cluster membersystems by entering the following command:# /sbin/sysconfig -r rm rm_use_512=

Table 5–2: MC2 Jumper Configuration (cont.)Jumper: Description: Example:J5: AlphaServer8x00 Mode8x00 mode selected:Pins 1 to 2a1238x00 mode not select

5.2 Installing the Memory Channel AdapterInstall the Memory Channel adapter in an appropriate peripheralcomponent interconnect (PCI) slot. (See Sectio

5.4 Installing the Memory Channel HubYou may use a hub in a two-node TruCluster Server cluster, but the hub isnot required. When there are more than t

______________________ Note _______________________Do not connect an MC1 or MC1.5 link cable to an MC2 module.5.5.1.1 Connecting MC1 or MC1.5 Link Cab

Figure 5–1 shows Memory Channel adapters connected to linecards that arein the same slot position in the Memory Channel hubs.Figure 5–1: Connecting Me

Do not connect an MC2 cable to an MC1 or MC1.5 CCMAAmodule.Gently push the cable’s connector into the receptacle, and then use thescrews to pull the c

5.5.2.4 Connecting MC2 Cables in Standard Hub Mode Using Optical ConvertersIf you are using optical converters in an MC2 configuration, install an opt

6. Install the CCMFB fiber-optic converter in slot opto only, 0/opto,1/opto, 2/opto,or3/opto, as appropriate.7. Install a BN39B-01 1-meter (3.3-foot)

When the console indicates a successful response from all other systemsbeing tested, the data flow through the Memory Channel hardware hasbeen complet

6–4Displaying the UDID and Worldwide Names of Devices Knownto the Console ...6–626–5Setting the

5.7 Maintaining Memory Channel InterconnectsThe following sections contain information about maintaining MemoryChannel interconnects. See other sectio

______________________ Note _______________________When you upgrade from dual, redundant MC1 hardware todual, redundant MC2 hardware, you must replace

Table 5–4: Adding a Memory Channel Interconnect or Upgrading from aDual, Redundant MC1 Interconnect to MC2 Interconnects (cont.)StepAction Refer to:MC

Table 5–4: Adding a Memory Channel Interconnect or Upgrading from aDual, Redundant MC1 Interconnect to MC2 Interconnects (cont.)StepAction Refer to:Vi

Table 5–4: Adding a Memory Channel Interconnect or Upgrading from aDual, Redundant MC1 Interconnect to MC2 Interconnects (cont.)StepAction Refer to:•

[2] 65536 4[3] 0[4] 655364[5] 0[6] 0[7] 0}(dbx) p rm_adapters[1]->rmp_prail_va->rmc_size5{[0] 163846[1] 0[2] 163846[3] 0[4] 163846[5] 0[6] 0[7]

Figure 5–2 shows a dual, redundant virtual hub configuration using MC1hardware being upgraded to MC2.Figure 5–2: MC1-to-MC2 Virtual Hub UpgradeAlphaSe

Figure 5–3 through Figure 5–8 show a three-node standard hubconfiguration being upgraded from MC1 to MC2.Figure 5–3: MC1-to-MC2 Standard Hub Upgrade:

Figure 5–4: MC1-to-MC2 Standard Hub Upgrade: First MC1 ModuleReplacedAlphaServerMemberSystem 1MC2MC1AlphaServerMemberSystem 2MC1MC1MC1 Hub# 1MC1 Hub#

Figure 5–5: MC1-to-MC2 Standard Hub Upgrade: Replace First MC1Adapter in Second SystemAlphaServerMemberSystem 1MC2MC1AlphaServerMemberSystem 2MC2MC1MC

2–1PCI Backplane Slot Layout ... ... ... ...2–33–1VHDCI Trilink Connector (H8861-AA) ...3–83–2D

Figure 5–6: MC1-to-MC2 Standard Hub Upgrade: Replace Third SystemMemory Channel AdaptersAlphaServerMemberSystem 1MC2MC1AlphaServerMemberSystem 2MC2MC1

Figure 5–7: MC1-to-MC2 Standard Hub Upgrade: Replace Second MC1 inSecond SystemAlphaServerMemberSystem 1MC2MC1AlphaServerMemberSystem 2MC2MC2MC2 Hub#

Figure 5–8: MC1-to-MC2 Standard Hub Upgrade: Final ConfigurationAlphaServerMemberSystem 1MC2MC2AlphaServerMemberSystem 2MC2MC2MC2 Hub# 1MC2 Hub# 2Alph

There will be some cluster down time. During the procedure, you canmaintain cluster operations except for the time it takes to shut down thesecond sys

Table 5–5: Upgrading from a Virtual Hub Configuration to a Standard HubConfiguration (cont.)StepAction Refer to:______________________ Note __________

Table 5–5: Upgrading from a Virtual Hub Configuration to a Standard HubConfiguration (cont.)StepAction Refer to:______________________ Note __________

6Using Fibre Channel StorageThis chapter provides an overview of Fibre Channel, Fibre Channelconfiguration examples, and information on Fibre Channel

______________________ Note _______________________TruCluster Server Version 5.1A configurations require one ormore disks to hold the Tru64 UNIX opera

AL_PAThe Arbitrated Loop Physical Address (AL_PA) isused to address nodes on the Fibre Channel loop.When a node is ready to transmit data, it transmit

6–8A Configuration That Is Not Recommended .. ... ... 6–156–9Another Configuration That Is Not Recommended ... ... ... .6–166–10Arbi

F_Port The ports within the fabric (fabric port). This port iscalled an F_port. Each F_port is assigned a 64-bitunique node name and a 64-bit unique p

6.1.2.1 Point-to-PointThe point-to-point topology is the simplest Fibre Channel topology. In apoint-to-point topology, one N_Port is connected to anot

Figure 6–2 shows an example fabric topology.Figure 6–2: Fabric TopologyN_PortNode 1Node 2TransmitReceiveTransmitReceiveTransmitReceiveTransmitReceiveT

Figure 6–3: Arbitrated Loop TopologyNL_PortNL_PortNode 1Node 2NL_PortNL_PortTransmitReceiveTransmitReceiveTransmitReceiveTransmitReceiveNode 3Node 4Hu

Although the fabric topology is more expensive, it provides both increasedconnectivity and higher performance; switches provide a full-duplex 100(200)

Figure 6–4 shows a typical Fibre Channel cluster configuration usingtransparent failover mode.Figure 6–4: Fibre Channel Single Switch Transparent Fail

Figure 6–5 shows a two-node Fibre Channel cluster with a single RA8000or ESA12000 storage array with dual-redundant HSG80 controllers and anDS-SWXHB-0

• Normally, all available units (D0 through D199) are available at allhost ports. Only one HSG80 controller will be actively doing I/O forany particul

Figure 6–6: Multiple-Bus NSPOF Configuration Number 1ZK-1707U-AIKGPSAMemberSystem2KGPSAMemoryChannelKGPSAMemberSystem1KGPSAMemoryChannelMemoryChannelM

Figure 6–7: Multiple-Bus NSPOF Configuration Number 2KGPSAMemberSystem2KGPSAKGPSAMemberSystem1KGPSAMemoryChannelMemoryChannelMemoryChannelMemoryChanne

10–10Two BA356s Cabled for Shared SCSI Bus Usage ... 10–2210–11Two UltraSCSI BA356s Cabled for Shared SCSI Bus Usage . . 10–2410–12Extern

The configuration that is shown in Figure 6–8 is a NSPOF configuration,but is not a recommended cluster configuration because of the performanceloss d

Figure 6–8: A Configuration That Is Not RecommendedZK-1706U-AIKGPSAMemberSystem2KGPSAMemoryChannelKGPSAMemberSystem1KGPSAMemoryChannelMemoryChannelMem

Figure 6–9: Another Configuration That Is Not RecommendedZK-1806U-AIKGPSAAlphaServerKGPSAFibre Channel SwitchFibre Channel SwitchHSG 80Controller AHSG

Figure 6–10 shows the maximum supported arbitrated loop configuration ofa two-node Fibre Channel cluster with two RA8000 or ESA12000 storagearrays, ea

private arbitrated loops (looplets) that are interconnected by a fabric. Aprivate loop is formed by logically connecting ports on up to two switches._

Switch zoning controls access at the storage system level, whereas SSPcontrols access at the storage unit level.The following configurations require z

If a host attempts to access a port that is outside its zone, the switchhardware blocks the access.You must modify the zone configuration when you mov

Figure 6–11: A Simple Zoned ConfigurationKGPSAMemoryChannelKGPSAKGPSAKGPSAFibre Channel SwitchCluster 1MemberSystem 1Cluster 1MemberSystem 2Cluster 2M

Figure 6–12: Meshed Fabric with Three Cascaded SwitchesKGPSAMemberSystem1MemoryChannelKGPSAMemoryChannelMemoryChannelInterfaceMemberSystem2HSG 80Contr

Figure 6–13: Meshed Resilient Fabric with Four Cascaded SwitchesKGPSAMemberSystem2KGPSAMemoryChannelMemory ChannelInterfaceHSG 80Controller AHSG 80Con

4–1Planning Your Configuration ... . ... ...4–44–2Configuring TruCluster Server Hardware ... ... ... 4–64–

______________________ Note _______________________If you lose an ISL, the communication can be routed throughanother switch to the same port on the o

6. Use the show wwid* and show n* console commands to show thedisk devices that are currently reachable, and the paths to the devices(Section 6.9.1.4)

6.8 Installing and Configuring Fibre Channel HardwareThis section provides information about installing the Fibre Channelhardware that is needed to su

DS-DSGGC-AA/AB) before you can manage the switch via atelnet session, SNMP, or the Web.The DS-DSGGC-AA/AB Fibre Channel switches have a default IPaddr

For an installation, at a minimum, you have to complete the following steps.Some of the steps are explained in more detail in the following sections.1

6.8.1.2 Managing the Fibre Channel SwitchesYou can manage the DS-DSGGA-AA, DS-DSGGA-AB, and DS-DSGGB-ABswitches, and obtain switch status from the fro

6.8.1.2.2 Setting the Ethernet IP Address and Subnet Mask from the Front PanelBefore you telnet to the switch, you must connect the Ethernet cable and

Press Enter to display the first submenu in the Operation Menu, SwitchOffline:Operation Menu:Switch Offline8. Press the down button until the Reboot s

4. Turn on power to the switch and log in. If the connection is correct, theself-test results will be displayed. It takes 2 to 3 minutes for self-test

Table 6–3: Telnet Session Default User Names for Fibre Channel SwitchesDSGGADSGGB orDSGGC Descriptionothern/a Allows you to execute commands ending in

8–19Hardware Components Used to Create the ConfigurationShown in Figure 8–18 ... ...8–678–20Hardware Compone

______________________ Note _______________________When you telnet to the switch the next time, the prompt willinclude the switch name, for example:fc

_____________________ Caution _____________________Static electricity can damage modules and electronic components.We recommend using a grounded antis

• Solid amber: Indicates that a loss of signal or poor signal integrity hasput the port in bypass mode. Make sure that a GBIC is installed, thata cabl

Remember to remove the transparent plastic covering on the extremitiesof the optical cable.5. Connect the fiber-optic cables to the shortwave Gigabit

P00>>> wwidmgr -show adapterwwidmgr available only prior to booting.Reinit system and try again.P00>>> init...P00>>> wwidmg

If the current topology for an adapter is LOOP, set an individual adapter toFABRIC by using the item number for that adapter (for example, 0 or 1).Use

fabric mode that is connected to a loop. Therefore, determine thetopology setting before using the adapter.The wwidmgr utility is documented in the Ww

6. Repeat this process for the other cluster member if this is a two-nodeTruCluster configuration.6.8.3.4 Obtaining the Worldwide Names of KGPSA Adapt

6.8.4 Setting Up the HSG80 Array Controller for Tru64 UNIXInstallationThis section covers setting up the HSG80 controller for operation with Tru64UNIX

6. Install disks into storage shelves.7. Connect a terminal to the maintenance port on one of the HSG80controllers. You need a local connection to con

About This ManualThis manual describes how to set up and maintain the hardwareconfiguration for a TruCluster™ Server cluster.AudienceThis manual is fo

HSG80> set other port_1_topology = offline6HSG80> set other port_2_topology = offline 6HSG80> set this port_1_topology = fabric 7HSG80> se

Setting the SCSI_VERSION to SCSI-2 allows a disk unit to be atLUN 0, and specifies that the command console LUN (CCL) is notfixed at a particular loca

Example 6–1: Determine HSG80 Connection Names (cont.)!NEWCON54 TRU64_UNIX OTHER 1 230813 OL other 0HOST_ID=1000-0000-C920-DA01 ADAPTER_ID=1000-0000-C9

HSG80 (!NEWCON50). For example, assume that membersystem pepicelli has two KGPSA Fibre Channel hostbus adapters, and that the worldwide name for KGPSA

4 Verifies that all connections have the offsets set to 0 and theoperating system is set to TRU64_UNIX.____________________ Note _____________________

This is the preferred address, but the HSG80 controller is free to usewhatever AL_PA it obtains during loop initialization. However, theaddress you sp

HSG80> set other PORT_2_AL_PA = 02After you have done this, continue with steps 12 through 14 in Section 6.8.4.6.8.4.2 Obtaining the Worldwide Name

controller, there are different procedures for replacing HSG80 controllersin an RA8000 or ESA12000:• If you replace one controller of a dual-redundant

This section describes how to perform the following tasks:• Before the installation:a. Configure HSG80 storagesets — In this manual, examplestorageset

f. Add additional systems to the cluster (Section 6.9.7).If you are installing either the Tru64 UNIX operating system or TruClusterServer software, fo

© 2001 Compaq Computer CorporationCompaq, the Compaq logo, AlphaServer, StorageWorks, and TruCluster Registered in U.S. Patent andTrademark Office. Al

Chapter 3 Contains information about setting up a shared SCSI bus, SCSI bus requirements, and how to connect storage to a shared SCSI bususing the lat

configuration. A blank table (Table A–1) is provided in Appendix A for use inan actual installation.One mirrorset, the BOOT-MIR mirrorset, is used for

Example 6–2: Setting Up the Mirrorset (cont.)HSG80> CREATE_PARTITION CROOT-MIR SIZE=56HSG80> CREATE_PARTITION CROOT-MIR SIZE=15 6HSG80> CREAT

3 Initializes the BOOT-MIR and CROOT-MIR mirrorsets. If you want to setany initialization switches, you must do so in this step. The BOOT-MIRmirrorset

______________________ Note _______________________A storageset must reside on one controller or the other. All thepartitions of a storageset must be

Example 6–3: Adding Units and Identifiers to the HSG80 Storagesets(cont.)HSG80> set d144 ENABLE_ACCESS_PATH = !NEWCON61,!NEWCON62,!NEWCON64,!NEWCON

selective access in case there are other systems or clusters that areconnected to the same switch as the cluster.Record the unit name of each partitio

HSG80. The identifiers should be easily recognized. Ensurethat you record the identifiers (Table 6–4).3Enables access to each unit for those hosts tha

Table 6–4: Converting Storageset Unit Numbers to Disk Names (cont.)File Systemor DiskHSG80Unit Worldwide Name UDID Device Name dskn/usrD143 6000-1FE1-

wwid3P00>>> show n*N1N2N3N4____________________ Note _____________________The console only creates devices for which the wwidnconsole environ

1 The number within the brackets ([ ]) is the item number of thedevice shown on any particular line.2 The UDID is assigned at the HSG80 with the set D

• Cluster Administration — Describes cluster-specific administrationtasks.• Cluster Highly Available Applications — Describes how to deployapplication

Example 6–5 shows:• The use of the wwidmgr -quickset command to set the device unitnumber for the Tru64 UNIX Version 5.1A installation disk to 133,and

Example 6–5: Setting the Device Unit Number with the wwidmgr quicksetCommand (cont.)via adapter: via fc nport: connected:dga131.1001.0.1.0 pga0.0.0.1.

6.9.1.4 Displaying the Available Boot DevicesThe only Fibre Channel devices that are displayed by the console show devcommand are those devices that h

Clear the wwid1 console environment variable as follows:P00>>> wwidmgr -clear wwid1Then, reboot the system.Example 6–6 provides sample device

______________________ Note _______________________The bootdef_dev environment variable values must point tothe same HSG80.To set the bootdef_dev cons

P00>>> show bootdef_devbootdef_dev dga133.1002.0.1.0You are now ready to install the Tru64 UNIX operating system.6.9.2 Install the Base Opera

# hwmgr -view dev | grep IDENTIFIERHWID: Device Name Mfg Model Location-----------------------------------------------------------------------62: /dev

...For more information on the hardware manager, seehwmgr(8).2. Search the display for the UDIDs (or worldwide names) for each of thecluster installat

6.9.6 Reset the bootdef_dev Console Environment VariableIf you set the bootdef_dev console environment variable to multiple pathsin Section 6.9.1.5, t

You can set units preferred to a specific controller, in whichcase both controllers will be active.If the bootdef_dev console environment variable end

• MA6000 HSG60 Array Controller ACS Version 8.5 Solution Software forCompaq Tru64 UNIX Installation and Configuration Guide• Compaq StorageWorks HSG60

______________________ Note _______________________The console System Reference Manual (SRM) software guaranteesthat you can set the bootdef_dev conso

c. Boot genvmunix on the newly added cluster member system. Eachinstalled subset will be configured and a new kernel will be builtand installed. After

4 Path from host bus adapter A to controller B port 2c. Set the bootdef_dev console environment variable for membersystem 2 boot disk to a comma-separ

connection is discovered. In transparent failover mode, host connectionsto port 1 default to an offset of 0; host connections on port 2 default to ano

2. At the HSG80, set multiple-bus failover as follows. Before putting thecontrollers in multiple-bus failover mode, you must remove any previousfailov

!NEWCON54 TRU64_UNIX OTHER 1 230813 OL other 0HOST_ID=1000-0000-C920-DA01 ADAPTER_ID=1000-0000-C920-DA01!NEWCON55 TRU64_UNIX OTHER 2 230913 OL other 1

____________________ Note _____________________The remaining steps apply only to fabric configurations. Inthis release, you cannot boot from storage t

the example, cluster member 1 will need access to the storage unitswith UDIDs 131 (member 1 boot disk) and 133 (Tru64 UNIX disk).Cluster member 2 will

• Display the current Fibre Channel topology for a Fibre Channel adapterSee emxmgr(8)for more information on the emxmgr utility.6.11.1 Using the emxmg

N_Port at FC DID 0x210113 - SCSI tgt id 1 : 2portname 5000-1FE1-0001-8931nodename 5000-1FE1-0001-8930Present, Logged in, FCP Target, FCP Logged in,N_P

• Compaq StorageWorks Heterogeneous Open SAN Design Reference Guide• Fibre Channel Storage Hub 7 Installation Guide• Fibre Channel Storage Hub 7 Rack

• FCP Initiator — The remote N_Port acts as a SCSI initiator device(it sends SCSI commands).• FCP Suspended — The driver has invoked a temporary suspe

Present, Logged in, FCP Target, FCP Logged in,N_Port at FC DID 0x00006e - SCSI tgt id 4 :portname 2200-0020-3700-55CBnodename 2000-0020-3700-55CBPrese

1. View adapter’s current Topology2. View adapter’s Target Id Mappings3. Change Target ID Mappingsd. Display Attached Adaptersa. Change targeted adapt

7Using GS80, GS160, or GS320 HardPartitions in a TruCluster ServerConfigurationThis chapter contains information about using AlphaServer GS80/160/320h

The AlphaServer GS80/160/320 systems use the same switch technology, thesame CPU, memory, and power modules, and the same I/O riser modules.The GS160

Figure 7–1: Portion of QBB Showing I/O Riser ModulesI/O RiserBN39B I/O Riser Cable ZK-1749U-AI____________________ Notes ____________________You can h

We recommend that you connect I/O riser 0 (local I/O riserports 0 and 1) to the primary PCI drawer that will be themaster system control manager (SCM)

1) that is higher than the master SCM. Both the master SCMand standby SCM must have the scm_csb_master_eligibleSCM environment variable set.__________

types of PCI drawers. It is harder to distinguish the type of PCI drawerfrom the rear, but slot 1 provides the key. The primary PCI drawer has astanda

Figure 7–3: Rear View of Expansion and Primary PCI DrawersZK-1751U-AIExpansion PCI DrawerPrimary PCI DrawerConsole SerialBus Node ID ModuleLocal Termi

– ESL9326 Tape Drive Upgrade GuideThe Golden Eggs Visual Configuration Guide provides configurationdiagrams of workstations, servers, storage componen

equally well with any number of partitions (as supported by the systemtype) by modifying the amount and placement of hardware and the SCMenvironment v

• Shared storage that is connected to KZPBA-CB (parallel SCSI) orKGPSA-CA (Fibre Channel) host bus adapters.• Network controllers.3. Install BN39B cab

____________________ Notes ____________________If the OCP key switch is in the On or Secure position, thesystem will go through the power-up sequence.

Example 7–1: Defining Hard Partitions with SCM Environment Variables(cont.)hp_qbb_mask5 0hp_qbb_mask6 0hp_qbb_mask7 0srom_mask ff fxsrom_mask ff ff ff

scm_csb_master_eligible environment variable. The master andstandby SCM must be connected to the OCP. The master SCM musthave the lowest node ID.Use t

1 Turns on power to partition 0.2 Turns on power to partition 1.3 Transfers control from the SCM firmware to the SRM consolefirmware._________________

7.4 Determining AlphaServer GS80/160/320 SystemConfigurationYou may be required to reconfigure an AlphaServer GS80/160/320 systemthat is not familiar

1 Hard partition number. There are two hard partitions in this example(0 and 1).2 QBB number and console serial bus (CSB) node ID. QBB 0 and 1 (CSBnod

11 Hierarchical switch (H-switch) type, status, temperature, and a reportof which QBBs are connected to the H-switch. In this example, QBBs0, 1, 2, an

Example 7–4: Displaying Console Serial Bus Information (cont.)30 PSM T05.4 (03.24/01:09) T4.0 (07.06) ON SrvSw: NORMAL30 XSROM T05.4 (03.24/02:10)C0 C

P Manuals for programmersR Manuals for reference page usersSome manuals in the documentation help meet the needs of severalaudiences. For example, the

• PBM (PCI backplane manager)• PSM (Power system manager)• HPM (Hierarchical switch power manager)• SCM master: This PCI primary drawer has the master

• System Reference Manual (SRM) flash ROM on the standard I/O module• The flash ROMs for the following console serial bus (CSB)microprocessors:– SCM:

____________________ Note _____________________You do not need to zero the hp_qbb_maskn environmentvariables, only the hp_count.5. Turn power on to th

Use the update command to update all firmware, or you can designatea specific device to update; for example, SRM console firmware:UPD> update srm__

8Configuring a Shared SCSI Bus for TapeDrive UseThe topics in this section provide information on preparing the various tapedevices for use on a share

They both work with an expansion unit (previously called theDS-TL890-NE) and a new module called the data unit.Section 8.12 covers the TL881 and TL891

Figure 8–1: TZ88N-VA SCSI ID SwitchesSnap−inLockingHandlesSCSI IDSwitch PackBackplaneInterfaceConnectorTZ88N−VATable 8–1: TZ88N-VA Switch SettingsSCSI

Table 8–1: TZ88N-VA Switch Settings (cont.)SCSI ID SCSI ID Selection Switches6Off On On Off Off Off7On On On Off Off OffaSBB tape drive SCSI ID is det

Figure 8–2 shows a TruCluster Server cluster with three shared SCSI buses.One shared bus has a BA350 with a TZ88N-VA at SCSI ID 3.Figure 8–2: Shared S

ConventionsThe following typographical conventions are used in this manual:#A number sign represents the superuser prompt.% catBoldface type in intera

Table 8–2: Hardware Components Used to Create the Configuration Shownin Figure 8–2 (cont.)Callout NumberDescription8DWZZA-VA with H885-AA trilink conn

The single-ended SCSI bus may be daisy chained from one single-endedtape drive to another with BC19J cables as long as the SCSI bus maximumlength is n

Figure 8–3: DS-TZ89N-VW SCSI ID SwitchesSnap−inLockingHandlesSCSI IDSwitch PackBackplaneInterfaceConnectorDS−TZ89N−VWThe SCSI ID is selected by switch

Table 8–3: DS-TZ89N-VW Switch Settings (cont.)SCSI ID SCSI ID Selection Switches123456785On Off On Off Off Off Off Off6Off On On Off Off Off Off Off7O

8.2.3 Setting the DS-TZ89N-TA SCSI IDThe DS-TZ89N-TA has a push-button counter switch on the rear panel toselect the SCSI ID. It is preset at the fact

8.3 Compaq 20/40 GB DLT Tape DriveThe Compaq 20/40 GB DLT Tape Drive is a Digital Linear Tape (DLT)tabletop cartridge tape drive that can hold up to 4

Figure 8–4: Compaq 20/40 GB DLT Tape Drive Rear Panel+-020/40 GB DLT Tape DriveSCSI IDSelectorSwitchSCSI ID+-0ZK-1603U-AI8.3.2 Cabling the Compaq 20/4

(65.6-foot) cable). Ensure that the trilink or Y cable at both ends ofthe differential segment of the shared SCSI bus is terminated with anHD68 differ

Figure 8–5: Cabling a Shared SCSI Bus with a Compaq 20/40 GB DLT TapeDriveKZPBA-CB (ID 7)MemoryChannelInterfaceMemory ChannelKZPBA-CB (ID 6)Memory Cha

Table 8–4: Hardware Components Used to Create the Configuration Shownin Figure 8–5 (cont.)Callout NumberDescription9199629-002 or 189636-002 (68-pin h

1IntroductionThis chapter introduces the TruCluster Server product and some basiccluster hardware configuration concepts.The chapter discusses the fol

8.4.2 Cabling the Compaq 40/80-GB DLT DriveThe Compaq 40/80-GB DLT Drive is connected to a single-ended segmentof the shared SCSI bus. Figure 8–6 show

Ensure that SCSI IDs for the tape drive and host bus adapter do not conflict.To achieve system performance capabilities, we recommend that you placeno

Table 8–5: Hardware Components in the Configuration in Figure 8–6 (cont.)Callout NumberDescription7328215-00X, BN21K, or BN21L HD68 to HD68 cablec8H88

4. Issue a bus reset or turn the minilibrary power off and on again to causethe drive to recognize the new SCSI ID.8.5.2 Cabling the TZ885 Tape DriveT

______________________ Note _______________________Ensure that there is no conflict with tape drive and host busadapter SCSI IDs.Figure 8–7: Cabling a

Table 8–6: Hardware Components Used to Create the Configuration Shownin Figure 8–7 (cont.)Callout NumberDescription8H885-AA trilink connector9BN21M ca

Figure 8–8: TZ887 DLT MiniLibrary Rear Panel+-0TZ887ZK-1461U-AISCSI IDSCSI IDSelectorSwitch+-08.6.2 Cabling the TZ887 Tape DriveThe TZ887 is connected

length is not exceeded and there are sufficient SCSI IDs available. Ensurethat the tape drive on the end of the bus is terminated with an H8574-A orH8

8.7 Preparing the TL891 and TL892 DLT MiniLibraries forShared SCSI Usage______________________ Note _______________________To achieve system performan

The first and second lines of the default screen show the status of the twodrives (if present). The third line shows the status of the library robotic

executing in a cluster. They can access their disk data from any memberin the cluster.• Like the TruCluster Production Server Software product, TruClu

4. Select the tape drive (DLT0 Bus ID: or DLT1 Bus ID:) or library robotics(LIB Bus ID:) whose SCSI bus ID you want to change. The defaultSCSI IDs are

SCSI bus without stopping all ASE services that generate activityon the bus.For this reason, we recommend that tape devices be placed onseparate share

To connect the drive robotics and one drive to one shared SCSI bus and thesecond drive to a second shared SCSI bus, follow these steps:1. Connect a BN

Figure 8–10: TruCluster Server Cluster with a TL892 on Two Shared SCSIBusesKZPBA-CB (ID 7)MemoryChannelInterfaceMemory ChannelKZPBA-CB (ID 6)Member Sy

8.8 Preparing the TL890 DLT MiniLibrary Expansion UnitThe topics in this section provide information on preparing the TL890 DLTMiniLibrary expansion u

8.8.2.1 Cabling the DLT MiniLibrariesYou must make the following connections to render the DLT MiniLibrarysystem operational:• Expansion unit to the m

____________________ Notes ____________________Do not connect a SCSI bus to the SCSI connectors for thelibrary connectors on the base modules.We recom

Figure 8–11: TL890 and TL892 DLT MiniLibraries on Shared SCSI BusesTL892TL892TL890LibraryRoboticsDLT1Expansion ModulesRoboticsControlcablesDLT2Library

Table 8–8: Hardware Components Used to Create the Configuration Shownin Figure 8–11Callout NumberDescription1BN38C or BN38D cablea2BN37A cableb3H8861-

After a series of power-on self-tests have executed, the default screenwill be displayed on the base module control panel:DLT0 IdleDLT1 IdleLoader Idl

• Optionally, one disk on a shared SCSI bus to act as the quorum disk (seeSection 1.3.1.4). For a more detailed discussion of the quorum disk, seetheC

inventory of modules may be incorrect and the contents of some or all of themodules will be inaccessible to the system and to the host.When the expans

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

8.9 Preparing the TL894 DLT Automated Tape Library forShared SCSI Bus UsageThe topics in this section provide information on preparing the TL894 DLTau

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 re

Menu:Configuration:4. Press and release SELECT to choose the Configuration menu.5. Verify that the following information is displayed in the SDA:Menu:

This configuration, which is called the four-bus configuration, is shown inFigure 8–12. In this configuration, each of the tape drives, except SCSI bu

Appendix B of the TL81X/TL894 Automated Tape Libraryfor DLT Cartridges Facilities Planning and Installation Guideprovides figures showing various bus

In Figure 8–13, one bus is connected to port 1 (robotics controller and tapedrives 0 and 1) and the other bus is connected to port 3 (tape drives 2 an

Table 8–10: Hardware Components Used to Create the ConfigurationShown in Figure 8–13 (cont.)Callout NumberDescription6H879-AA terminator7328215-00X, B

is applied, until the electronics is activated and able to set theSCSI IDs electronically.The physical SCSI IDs should match the SCSI IDs set by theli

ContentsAbout This Manual1 Introduction1.1The TruCluster Server Product ... . ... ... 1–11.2Memory Requirements . ...

file system cannot also be used as the member boot disk or as the quorumdisk.1.3.1.3 Member Boot DiskEach member has a boot disk. A boot disk contains

2. On the Enter Password screen, enter the operator password. Thedefault operator password is 1234. The lock icon is unlocked and showsan O to indicat

In this configuration each of the tape drives, except tape drive 0 and therobotics controller, require a SCSI ID on a separate SCSI bus. The roboticsc

Figure 8–14: TL895 Tape Library Internal CablingRoboticsControllerSCSI ID 0SCSI Port 5SCSI Port 4SCSI Port 3 Tape Drive 0SCSI ID 1 Tape Drive 1SCSI

electronic SCSI ID using the Configure menu from the control panel (seeSection 8.10.2).The actual upgrade is beyond the scope of this manual. See the

Each tape library comes configured with a robotic controller and bar codereader (to obtain quick and accurate tape inventories).The libraries have eit

These tape libraries each have a multi-unit controller (MUC) that servestwo functions:• It is a SCSI adapter that allows the SCSI interface to control

Table 8–12: MUC Switch Functions (cont.)SwitchFunction7Host selection: Down for SCSI, up for seriala8Must be down, reserved for testingaFor a TruClust

Table 8–15: TL896 Default SCSI IDsSCSI PortDeviceDefault SCSI IDMUC2DDrive 5 (top)5E Drive 44F Drive 33A Drive 25B Drive 14C Drive 0 (bottom)38.11.5 T

Figure 8–15: TL893 Three-Bus ConfigurationMUCSCSI Address 2SCSI Port A SCSI Port B SCSI Port CTZ89 Tape DriveSCSI Address 5 (top shelf)TZ89 Tape Dr

– The lower bay bottom shelf tape drive (tape drive 0, SCSI ID 3) is onSCSI Port C and is terminated on the tape drive.– The tape drive terminators ar

• A quorum disk can have either 1 vote or no votes. In general, a quorumdisk should always be assigned a vote. You might assign an existingquorum disk

other devices on the shared SCSI bus. Each SCSI bus must be terminatedinternal to the tape library at the tape drive itself with the installed SCSIter

Figure 8–17: Shared SCSI Buses with TL896 in Three-Bus ModeKZPBA-CB (ID 7)MemoryChannelInterfaceMemory ChannelKZPBA-CB (ID 6)DS-DWZZH-03TTT21413Storag

Table 8–16: Hardware Components Used to Create the ConfigurationShown in Figure 8–17 (cont.)Callout NumberDescription6H879-AA terminator7328215-00X, B

8.12.1.2 TL881 and TL891 MiniLibrary Rackmount ComponentsA TL881 or TL891 base unit (which contains the tape drive) can operate asan independent, stan

• Data unit — This rackmount module contains a 16-cartridge magazineto provide additional capacity in a multi-module configuration. The dataunit robot

8.12.1.4 DLT MiniLibrary Part NumbersTable 8–18 lists the part numbers for the TL881 and TL891 DLT MiniLibrarysystems. Part numbers are only shown for

8.12.2.1 Preparing a Tabletop Model or Base Unit for Standalone Shared SCSIBus UsageA TL881 or TL891 DLT MiniLibrary tabletop model or a rackmount bas

______________________ Note _______________________There are no switches for setting a mechanical SCSI ID for thetape drives. The SCSI IDs default to

5. Use the up or down arrow button to scroll through the possible SCSI IDsettings. Press the Enter button when the desired SCSI ID is displayed.6. Rep

______________________ Note _______________________It is assumed that the library robotics controller is on the sameSCSI bus as tape drive 1.To connec

Figure 1–1: Two-Node Cluster with Minimum Disk Configuration and NoQuorum DiskMemberSystem1NetworkMemory ChannelShared SCSI BusPCI SCSIAdapterMemberSy

4. Install an HD68 differential (H879-AA) terminator on the right DLT1connector (the fourth connector from the left) and install another HD68different

Table 8–19: Hardware Components Used to Create the ConfigurationShown in Figure 8–18Callout NumberDescription1BN38C or BN38D cablea2BN37A cableb3H8861

_____________________ Note _____________________This cable is not shown in Figure 8–19 because thepass-through mechanism is not shown in the figure.•

pass-through mechanism and cable to the library robotics motor are notshown in this figure.For more information on cabling the units, see Section 8.12

Table 8–20: Hardware Components Used to Create the ConfigurationShown in Figure 8–19Callout NumberDescription1BN38C or BN38D cablea2BN37A cableb3H8861

DLT0 IdleDLT1 IdleLoader Idle0>__________<9The default screen shows the state of the tape drives, loader, andnumber of cartridges present for th

When the expansion unit comes up, it will communicate with each base anddata unit through the expansion unit interface and inventory the numberof base

• DLT3 Bus ID: 4• DLT4 Bus ID: 5• DLT5 Bus ID: 67. Press Enter when you have the item selected for which you want tochange the SCSI ID.8. Use the up a

These tape devices have been qualified for use on shared SCSI buseswith both the KZPSA-BB and KZPBA-CB host bus adapters. Ensurethat the host bus adap

8.13.3 Preparing the ESL9326D Enterprise Library for Shared SCSIBus UsageThe ESL9326D Enterprise Library contains library electronics (roboticcontroll

Figure 1–2: Generic Two-Node Cluster with Minimum Disk Configurationand Quorum DiskMemberSystem1NetworkMemory ChannelShared SCSI BusPCI SCSIAdapterMem

8.13.3.3 ESL9326D Enterprise Library Internal CablingThe default internal cabling for the ESL9326D Enterprise Library is to placetwo 35/70 DLT tape dr

______________________ Note _______________________Each internal cable is up to 2.5 meters (8.2 feet) long. The lengthof the internal cables, two per

Table 8–21: Shared SCSI Bus Cable and Terminator Connections for theESL9326D Enterprise LibraryTape Drives on SharedSCSI BusConnect SCSI Cableto Conne

9Configuring Systems for ExternalTermination or Radial Connections toNon-UltraSCSI DevicesThis chapter describes how to prepare the systems for a TruC

Channel adapters, hubs (if necessary), cables, and network adapters) havebeen installed, you can connect your host bus adapter to the UltraSCSIhub or

9.1.1 Radial Installation of a KZPSA-BB or KZPBA-CB Using InternalTerminationUse this method of cabling member systems and shared storage in aTruClust

The other end of the SCSI bus segment is terminated by the KZPSA-BBor KZPBA-CB onboard termination resistor SIPs, or a trilinkconnector/terminator com

Table 9–2: Installing the KZPSA-BB or KZPBA-CB for Radial Connectionto a DWZZH UltraSCSI HubStepAction Refer to:1Ensure that the KZPSA-BB internal ter

Table 9–2: Installing the KZPSA-BB or KZPBA-CB for Radial Connectionto a DWZZH UltraSCSI Hub (cont.)StepAction Refer to:5Use the show config and show

Table 9–3: Installing a KZPSA-BB or KZPBA-CB for Use with ExternalTerminationStepAction Refer to:1Remove the KZPSA-BB internal terminationresistors, Z

• Using a redundant array of independent disks (RAID) array controller intransparent failover mode allows the use of hardware RAID to mirror thedisks.

Table 9–3: Installing a KZPSA-BB or KZPBA-CB for Use with ExternalTermination (cont.)StepAction Refer to:7Use the show pk* or show isp* consolecommand

Table 9–3: Installing a KZPSA-BB or KZPBA-CB for Use with ExternalTermination (cont.)StepAction Refer to:TL891/TL892 MiniLibrary Section 8.7TL890 with

Example 9–1: Displaying Configuration on an AlphaServer 4100 (cont.)Memory 64 MB SYNC 0 0000 mem2Memory 64 MB SYNC 0 0000 mem3CPU (4MB Cache) 3 0000 c

Example 9–2: Displaying Devices on an AlphaServer 4100 (cont.)dkd2.0.0.4.1 DKd2 HSZ50-AX X29Zdkd100.1.0.4.1 DKd100 RZ26N 0568dkd200.1.0.4.1 DKd200 RZ2

Example 9–4: Displaying Devices on an AlphaServer 8200>>> show devicepolling for units on isp0, slot0, bus0, hose0...polling for units on isp

9.1.4 Displaying Console Environment Variables and Setting theKZPSA-BB and KZPBA-CB SCSI IDThe following sections show how to use the show console com

Example 9–5: Displaying the pk* Console Environment Variables on anAlphaServer 4100 System (cont.)pkf0_fast 1pkf0_host_id 7pkf0_termpwr 1Compare the s

Example 9–6: Displaying Console Variables for a KZPBA-CB on anAlphaServer 8x00 SystemP00>>> show isp*isp0_host_id 7isp0_soft_term onisp1_host

Example 9–7: Displaying Console Variables for a KZPSA-BB on anAlphaServer 8x00 System (cont.)pkc0_fast 1pkc0_host_id 7pkc0_termpwr on9.1.4.2 Setting t

9.1.4.3 Setting KZPSA-BB SCSI Bus ID, Bus Speed, and Termination PowerIf the KZPSA-BB SCSI ID is not correct, or if it was reset to 7 by thefirmware u

Figure 1–3: Minimum Two-Node Cluster with UltraSCSI BA356 Storage UnitID 5PWRSharedSCSIBusClusterwide/, /usr, /varMember 1Boot DiskMember 2Boot DiskQu

9.1.4.4 KZPSA-BB and KZPBA-CB Termination ResistorsThe KZPSA-BB internal termination is disabled by removing terminationresistors Z1 through Z5, as sh

The boot sequence provides firmware update overview information. UseReturn to scroll the text, or press Ctrl/C to skip the text.After the overview inf

10Configurations Using ExternalTermination or Radial Connectionsto Non-UltraSCSI DevicesThis chapter describes the requirements for the shared SCSI bu

Introductory information covering SCSI bus configuration concepts (SCSIbus speed, data path, and so on) and SCSI bus configuration requirementscan be

host bus adapter to a non-UltraSCSI BA356 (single-ended and wide) storageshelf. The DS-BA35X-DA personality module is used in an UltraSCSI BA356to con

Figure 10–1: Standalone SCSI Signal ConverterZK-1050U-AITTSingle-ended Differential side with trilinkattachedsideFigure 10–2 shows the status of inter

______________________ Notes ______________________S4-3 and S4-4 have no function on the DS-BA35X-DA personalitymodule.See Section 10.3.2.2 for inform

Whenever possible, connect devices to a shared bus so that they can beisolated from the bus. This allows you to disconnect devices from the busfor mai

connector at a later time without affecting bus termination. This allows youto expand your configuration without shutting down the cluster.Figure 10–4

this slot can be used for a second power supply to provide fullyredundant power to the storage shelf.With the use of the cluster file system (see the

Figure 10–5: HD68 Trilink Connector (H885-AA)FRONT VIEWREAR VIEWZK-1140U-AI______________________ Note _______________________If you connect a trilink

10.3.1 BA350 Storage ShelfUp to seven narrow (8-bit) single-ended StorageWorks building blocks(SBBs) can be installed in the BA350. Their SCSI IDs are

Figure 10–6: BA350 Internal SCSI BusJA1JB10123456POWER (7)TJZK-1338U-AI10.3.2 BA356 Storage ShelfThere are two variations of the BA356 used in TruClus

select SCSI IDs 0 through 6, set the personality module address switches 1through 7 to off. To select SCSI IDs 8 through 14, set personality moduleadd

Figure 10–7: BA356 Internal SCSI BusJA1JB10123456POWER (7)JZK-1339U-AIJA1 and JB1 are located on the personality module (in the top of the boxwhen it

Figure 10–8: BA356 Jumper and Terminator Module Identification PinsSlot 6JumperPinSlot 6TerminatorPinSlot 1JumperPinSlot 1TerminatorPinZK-1529U-AI10.3

non-UltraSCSI BA356, as shown in Figure 10–8. With proper lighting youwill be able to see a J or T near the hole where the pin sticks through.Terminat

10.4.1 Preparing BA350, BA356, and UltraSCSI BA356 StorageShelves for an Externally Terminated TruCluster ServerConfigurationYou may be using the BA35

Remove the termination from the differential end by removing the five14-pin differential terminator resistor SIPs.3. Attach an H885-AA trilink connect

If you are using a DWZZB-VW, install it in slot 0 of the BA356.10.4.1.3 Preparing an UltraSCSI BA356 Storage Shelf for a TruCluster ConfigurationAn Ul

1.5.2 Two-Node Clusters Using UltraSCSI BA356 Storage Units withIncreased Disk ConfigurationsThe configuration shown in Figure 1–3 is a minimal config

10.4.2.1 Cabling a Single BA350 Storage ShelfTo cable a single BA350 storage shelf into a cluster, install a BN21K, BN21L,or 328215-00X HD68 cable bet

10.4.3.1 Connecting a BA350 and a BA356 for Shared SCSI Bus UsageWhen you use a BA350 and a BA356 for storage on a shared SCSI bus in aTruCluster Serv

Figure 10–9 shows a two-member TruCluster Server configuration usinga BA350 and a BA356 for storage.Figure 10–9: BA350 and BA356 Cabled for Shared SCS

Table 10–1: Hardware Components Used for Configuration Shown inFigure 10–9 and Figure 10–10Callout NumberDescription1BN21W-0B Y cable2H879-AA terminat

Figure 10–10 shows a two member TruCluster Server configuration usingtwo BA356s for storage.Figure 10–10: Two BA356s Cabled for Shared SCSI Bus UsageI

10.4.3.3 Connecting Two UltraSCSI BA356s for Shared SCSI Bus UsageWhen you use two UltraSCSI BA356 storage shelves on a shared SCSI busin a TruCluster

Figure 10–11 shows a two member TruCluster Server configuration usingtwo UltraSCSI BA356s for storage.Figure 10–11: Two UltraSCSI BA356s Cabled for Sh

Table 10–2: Hardware Components Used for Configuration Shown inFigure 10–11Callout NumberDescription1BN21W-0B Y cable2H879-AA HD68 terminator3BN38C (o

10.4.4.1 Cabling an HSZ40 or HSZ50 in a Cluster Using External TerminationTo connect an HSZ40 or HSZ50 controller to an externally terminated sharedSC

Figure 10–12 shows two AlphaServer systems in a TruCluster Serverconfiguration with dual-redundant HSZ50 RAID controllers in the middleof the shared S

Figure 1–4: Two-Node Cluster with Two UltraSCSI DS-BA356 Storage UnitsID 8ID 9ID 10ID 11ID 12ID 13PWRID 5PWRSharedSCSIBusClusterwide/, /usr, /varMembe

Figure 10–13: Externally Terminated Shared SCSI Bus with HSZ50 RAIDArray Controllers at Bus EndKZPSA-BB (ID 7)NetworkMemoryChannelInterfaceMemory Chan

10.4.4.2 Cabling an HSZ20 in a Cluster Using External TerminationTo connect a SWXRA-Z1 (HSZ20 controller) to a shared SCSI bus, followthese steps:1. R

______________________ Note _______________________The RA3000 is supported on a shared SCSI bus only with theKZPBA-CB UltraSCSI host bus adapter.Table

Table 10–4: Installing Cables for RA3000 Configuration Using ExternalTermination and Y Cables (cont.)Action Refer to:Install a BN37A-0E 50-centimeter

Figure 10–15: Externally Terminated TruCluster Server Configuration withan RA3000 Controller Shelf with Active/Passive FailoverAlphaServerMemberKZPBA-

Table 10–5: Hardware Components Used in the TruCluster ServerConfiguration Shown in Figure 10–14, Figure 10–15, and Figure 10–16Callout NumberDescript

Table 10–6: Hardware Components Used in the Configuration Shown inFigure 10–17Callout NumberDescription1H879-AA terminator2BN21W-0B Y cable3BN38C HD68

6. If you are using a:• DS-DWZZH-03: Install a BN38C (or BN38D) HD to VHDCI cablebetween any DS-DWZZH-03 port and the open connector on theH885-AA tri

Figure 10–18: TruCluster Server Cluster Using DS-DWZZH-03, SCSIAdapter with Terminators Installed, and HSZ50KZPSA-BB (ID 7)NetworkMemoryChannelInterfa

Figure 10–19: TruCluster Server Cluster Using KZPSA-BB SCSI Adapters,a DS-DWZZH-05 UltraSCSI Hub, and an HSZ50 RAID Array ControllerKZPSA-BB (ID 5)Net

______________________ Note _______________________You cannot use LSM to mirror the member system boot or quorumdisks, but you can use hardware RAID.F

11Configuring an Eight-Member ClusterUsing Externally Terminated Shared SCSIBusesThis chapter discusses the following topics:• Overview of an eight-no

The primary focus of this chapter is on an eight-node cluster that usesexternally terminated shared SCSI buses with minimal storage. This typeof clust

Figure 11–1: Block Diagram of an Eight-Node Cluster//usr/varMember 1,Member 2Boot DisksMember 6,Member 7,Member 8Boot DisksMember 3,Member 4,Member 5B

The Tru64 UNIX Version 5.1A operating system is installed onmember system 1. It can be installed on an internal disk, as is thecase in Figure 11–1, or

11.2 Configuring an Eight-Node Cluster Using an UltraSCSIBA356 and External TerminationConfiguring an eight-node cluster is carried out in three disti

1, which enables failover pair. See the Cluster Highly AvailableApplications manual for more information.Figure 11–2 provides a detailed illustration

7. Prepare the UltraSCSI BA356 for TruCluster Server use (seeSection 10.4.1.3). Ensure that you have installed an H8861-AA VHDCItrilink connector on t

Figure 11–2: First Two Nodes of an Eight-Node ClusterPWR/, /usr, /varMember 1Boot DiskMember 2Boot DiskUltraSCSI BA356KZPBA-CB (ID 6)MemberSystem

Table 11–1: Hardware Components Used for Configuration Shown inFigure 11–2Callout NumberDescription1BN21W-0B HD68 Y cable2H879-AA HD68 terminator3BN38

2.5RAID Array Controller Restrictions .. ...2–102.6SCSI Signal Converters ...2–12

use LSM to mirror the quorum or the member system boot disks, we do nothave a no-single-point-of-failure (NSPOF) cluster.1.5.4 Using Hardware RAID to

____________________ Note _____________________If member systems 1 and 2 are running cluster software,do not run mc_cable Memory Channel diagnostics.

9. Connect a BN21K, BN21L, or 328215-00X cable between the BN21W-0BY cables on member system 4 and member system 5.10. Connect a BN38C, BN38D, or a co

Table 11–2: Hardware Components Used for Configuration Shown Figure11–3Callout NumberDescription1BN21W-0B HD68 Y cable2H879-AA HD68 terminator3BN21K,

2. Use BN39B-04 (4 meters; 13.1 feet) or BN39B-10 (10 meters; 32.8 feet)to connect the Memory Channel adapters of member systems 6, 7, and 8to the Mem

Figure 11–4: Third Shared SCSI Bus of an Eight-Node ClusterMemberSystem 1TT132Memory ChannelKZPBA-CB (ID 7)2KZPBA-CB (ID 7)MemberSystem 6KZP

Table 11–3: Hardware Components Used for Configuration Shown inFigure 11–4Callout NumberDescription1BN21W-0B HD68 Y cable2H879-AA HD68 terminator3BN21

AWorldwide ID-to-Disk Name ConversionTableTable A–1: Converting Storageset Unit Numbers to Disk NamesFile Systemor DiskHSG80Unit WWID UDID Device Name

IndexNumbers and SpecialCharacters20/40-GB DLT Tape Drive, 8–11cabling, 8–12capacity, 8–11cartridges, 8–11connectors, 8–11setting SCSI ID, 8–1140/80-G

devices. Either controller can continue to service all of the units if the othercontroller fails.______________________ Note _______________________Th

extending differential, 10–2narrow data path, 3–5speed, 3–5terminating, 3–7, 10–5, 10–8wide data path, 3–5Ccaa_relocate command, 5–15t,5–27tcable leng

CSB, 7–4nodes, 7–5purpose, 7–4Ddata path for buses, 3–5default SCSI IDsESL9326D enterprise library, 8–75TL881/TL891 DLT MiniLibrary,8–63TL890 tape lib

SBB, 3–10SCSI ID, 3–10termpwr, 3–9transfer rate, 2–13DS-TZ89N-TA tabletop tape drivecabling, 8–10setting SCSI ID, 8–10DS-TZ89N-VW SBB tape drivecablin

FCP, 6–2ffauto console environmentvariable, 6–73ffnext console environmentvariable, 6–73fiber-optic cableFibre Channel, 2–8, 6–27Memory Channel, 2–5,

trilink connectors, 2–16hardware configurationbus termination, 3–7, 10–5cables supported, 2–1disk devices, 3–16, 10–14hardware requirements, 2–1hardwa

II/O busesnumber of, 2–6I/O riserscables, 7–3, 7–9local, 7–3, 7–9remote, 7–3, 7–9init command, 6–37, 6–65, 6–68,6–73, 6–76initializeafter setting boot

( See LFU )Logical Storage Manager( See LSM mirroring )loop topologyAL_PA, 6–3characteristics, 6–6defined, 6–3setting controller values, 6–48setting p

setting, 6–44, 6–78NN_Port node port, 6–3NL_Port node loop port, 6–3no single point of failure( See NSPOF )node name, 6–50non-Ultra BA356 storage shel

RAID, 1–14RAID Array 3000( See RA3000 )RAID array controllersadvantages, 3–17preparing, 10–25shared SCSI bus and, 10–25using in ASE, 10–25Redundant Ar

HSZ50 controller, 10–25in BA356, 10–11in UltraSCSI BA356, 10–13KZPBA-CB, 4–17RAID subsystem controllers, 10–25requirement, 3–6setting, 4–9t, 4–17, 8–2

____________________ Notes ____________________Only the HSZ70, HSZ80, HSG60, and HSG80 are capable ofsupporting multiple-bus failover (SET MULTIBUS_FA

UltraSCSI BA356 storage shelf,10–15, 10–17show config command, 4–10,4–13, 9–6t, 9–7t, 9–9, 9–11show csb command, 7–16show device command, 4–9t, 4–12,4

UltraSCSI BA356, 10–14termination resistorsKZPBA-CB, 4–9t, 4–17, 9–5t,9–7tKZPSA-BB, 9–5t, 9–7tterminators supported, 2–16TL881 tape library, 8–58TL881

personality module addressswitches, 10–13power supply, 3–3preparing, 10–15, 10–17preparing for shared SCSI usage,10–17SCSI ID selection, 10–13, 10–17t

Figure 1–8: NSPOF Fibre Channel Cluster Using HSG80s in Multiple-BusFailover ModeKGPSAMemberSystem2KGPSAKGPSAMemberSystem1KGPSAMemoryChannelMemoryChan

Figure 1–9 shows a two-member cluster configuration with three sharedSCSI buses. The clusterwide root (/), /usr, and /var file systems aremirrored acr

Figure 1–9: NSPOF Cluster Using LSM and UltraSCSI BA356sHost Bus Adapter (ID 7)MemoryChannelInterfaceMemory ChannelMemory ChannelMember System 2Member

1.6 Eight-Member ClustersTruCluster Server Version 5.1A supports eight-member clusterconfigurations as follows:• Fibre Channel: Eight-member systems m

For a Fibre Channel configuration, connect the HSG60 or HSG80controllers to the switches. You want the HSG60 or HSG80 to recognizethe connections to t

2Hardware Requirements and RestrictionsThis chapter describes the hardware requirements and restrictions fora TruCluster Server cluster. It includes l

4 TruCluster Server System Configuration Using UltraSCSIHardware4.1Planning Your TruCluster Server Hardware Configuration ... 4–24.2Obtaining the Firm

behind an HSZ80, HSG60, or HSG80 controller. If the cluster memberis using earlier firmware, the member may fail to boot, indicating"Reservation

Figure 2–1: PCI Backplane Slot LayoutI/O Riser 0I/O Riser 10-0/10-20-30-40-50-60-71-11-21-31-41-51-61-7PCI 0PCI 1PCI 1PCI 01-R 0-RZK-1748U-AI• TruClus

When the Memory Channel is set up in standard hub mode, the MemoryChannel hub must be visible to each member’s Memory Channel adapter.If the hub is po

• The maximum length of an MC2 BN39B link cable is 10 meters (32.8feet).• In an MC2 configuration, you can use a CCMFB optical converter inconjunction

Order an H3095-AA module to upgrade an AlphaServer 2000 or anH3096-AA module to upgrade an AlphaServer 2100 to support MemoryChannel.• For AlphaServer

Table 2–1: AlphaServer Systems Supported for Fibre Channel (cont.)AlphaServer SystemNumber of Adapters Supportedin Fabric TopologyNumber of AdaptersSu

Gigabit Interface Converter (GBIC) transceiver-based port connectionsfor maximum application flexibility. The hub is hot pluggable and isunmanaged.• O

• A storage array with dual-redundant HSG60 or HSG80 controllers inmultiple-bus failover is four targets and consumes four ports on a switch.• The HSG

>>> set bus_probe_algorithm newUse the show bus_probe_algorithm console command to determine ifyour system supports the variable. If the resp

Table 2–2: RAID Controller Minimum Required Array Controller SoftwareRAID ControllerMinimum Required Array ControllerSoftwareHSZ203.4HSZ22 (RAID Array

5.7.2Upgrading Memory Channel Adapters . ... ... 5–145.7.3Upgrading a Virtual Hub Configuration to a StandardHub Configuration ...

• Only RA3000 storage units visible to the host as LUN0 (storage unitswith a zero (0) as the last digit of the unit number such as D0, D100,D200, and

• If you observe any “bus hung” messages, your DWZZA signal convertersmay have the incorrect hardware. In addition, some DWZZA signalconverters that a

______________________ Note _______________________The DS-DWZZH-03/05 UltraSCSI hubs cannot be connected to aStorageWorks BA35X storage shelf because

Table 2–4: Supported SCSI Cables (cont.)CableConnectorDensity Pins Configuration UseBN38E-0B Technologyadapter cableHD68male toVHDCIfemaleMay be conne

2.9 SCSI Terminators and Trilink ConnectorsTable 2–5 describes the supported trilink connectors and SCSI terminatorsand the context in which you use t

3Shared SCSI Bus Requirements andConfigurations Using UltraSCSI HardwareA TruCluster Server cluster uses shared SCSI buses, external storageshelves or

______________________ Note _______________________Although the UltraSCSI BA356 might have been includedin this chapter with the other UltraSCSI devic

• Be careful when performing maintenance on any device that is on ashared bus because of the constant activity on the bus. Usually, toperform maintena

3.2 SCSI Bus PerformanceBefore you set up a SCSI bus, it is important that you understand a numberof issues that affect the viability of a bus and how

This transmission method is less susceptible to noise than single-endedSCSI and enables you to use longer cables. Devices with differentialSCSI interf

6.8.3.1Installing the KGPSA PCI-to-Fibre Channel AdapterModule ... ... ... ...6–366.8.3.2Setting the KGPSA

Table 3–1: SCSI Bus SpeedsSCSI Bus Transfer Rate (MHz) Bus Width (Bytes)Bus Bandwidth(Speed) MB/secSCSI515Fast SCSI10 1 10Fast-Wide10 2 20UltraSCSI20

3.4 SCSI Bus LengthThere is a limit to the length of the cables in a shared SCSI bus. The totalcable length for a SCSI bus segment is calculated from

______________________ Notes ______________________With the exception of the TZ885, TZ887, TL890, TL891, andTL892, tape devices can only be installed

of the bus segment. Termination for the other end of the bus segment isprovided by the folowing components:• Installed KZPBA-CB (or KZPSA-BB) terminat

– A StorageWorks UltraSCSI BA356 storage shelf (which has therequired 180-watt power supply).– The lower righthand device slot of the BA370 shelf with

The following section describes how to prepare the DS-DWZZH-05 UltraSCSIhub for use on a shared SCSI bus in more detail.3.6.1.2.1 DS-DWZZH-05 Configur

Table 3–3: DS-DWZZH UltraSCSI Hub Maximum ConfigurationsDS-DWZZH-03 DS-DWZZH-05 Disk DrivesaPersonalityModulebc500Not Installed400Installed303Installe

After the host adapter has been serviced, if there are still SCSI IDs retainedfrom the previous arbitration cycle, the next highest SCSI ID is service

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. TheDS-DWZZH-05 r

Figure 3–4: DS-DWZZH-05 Front ViewFair DisableHost PortSCSI ID1(13)Host PortSCSI ID3(15)ControllerPortSCSI ID6 - 4(6 - 0)Busy02ZK-1447U-AIHost PortHos

7 Using GS80, GS160, or GS320 Hard Partitions in a TruClusterServer Configuration7.1Overview . . ... ...

3.6.1.3 Installing the DS-DWZZH-05 UltraSCSI HubTo install the DS-DWZZH-05 UltraSCSI hub, follow these steps:1. Remove the W1 jumper to enable wide ad

the member system and storage device. Be aware though, theKZPSA-BB is not an UltraSCSI device and therefore only worksat fast-wide speed (20 MB/sec).T

Transparent failover compensates only for a controller failure, and notfor failures of either the SCSI bus or host adapters and is therefore not aNSPO

see their non-active host ports as passive. If one of the controllers fails, thesurviving controller sees both host ports as active.In the active/pass

• HSZ70 controller A and controller B• HSZ80 controller A Port 1 (2) and controller B Port 1 (2)The BN37A-0C is a 30-centimeter (11.8-inch) cable and

Figure 3–5: Shared SCSI Bus with HSZ70 Configured for TransparentFailoverKZPBA-CB (ID 7)NetworkMemoryChannelInterfaceMemory ChannelKZPBA-CB (ID 6)Memo

Table 3–4: Hardware Components Shown in Figure 3–5 Through Figure 3–8Callout NumberDescription1BN38C cablea2BN37A cableb3H8861-AA VHDCI trilink connec

3.7.1.2 Preparing a Dual-Redundant HSZ70 or HSZ80 for a Shared SCSI Bus UsingMultiple-Bus FailoverMultiple-bus failover is a dual-redundant controller

2. Install H8861-AA VHDCI trilink connectors (with terminators) on:• HSZ70 controller A and controller B• HSZ80 controller A Port 1 (2) and controller

Figure 3–7 shows a two-member TruCluster Server configuration witha radially connected dual-redundant HSZ70 configured for multiple-busfailover.Figure

8.8Preparing the TL890 DLT MiniLibrary Expansion Unit ... ... . 8–308.8.1TL890 DLT MiniLibrary Expansion Unit Hardware ... ... 8–308.8.2Preparing the

Figure 3–8 shows a two-member TruCluster Server configuration witha radially connected dual-redundant HSZ80 configured for multiple-busfailover.Figure

The RA3000 storage subsystem has fully redundant components to eliminatesingle points of failure. It comes with a standard uninterruptible powersupply

Table 3–5: Installing Cables for RA3000 Radial Configuration with aDWZZH UltraSCSI HubAction Refer to:Install a BN38C HD68 to VHDCI cable between each

Figure 3–9 shows a four-member TruCluster Server configuration and anRA3000 controller shelf with active/passive failover radially connected to aDS-DW

Figure 3–10 shows a four-member TruCluster Server configuration andan RA3000 pedestal with active/passive failover radially connected to aDS-DWZZH-05

Figure 3–11 shows a two-member TruCluster Server configuration andan RA3000 pedestal with active/active or active/passive failover radiallyconnected t

Figure 3–12 shows a four-member TruCluster Server configuration and anRA3000 controller shelf with active/active or active/passive failover radiallyco

Table 3–6: Hardware Components Used in the Configurations Shown inFigure 3–9 through Figure 3–12Callout NumberDescription1BN38C HD68-to-VHDCI cable.aA

4TruCluster Server System ConfigurationUsing UltraSCSI HardwareThis chapter describes how to prepare systems for a TruCluster Servercluster, using Ult