Hp 3PAR Operating System Software Uživatelský manuál

HP 3PAR Remote Copy Software User GuideHP 3PAR OS 3.1.3AbstractThis guide is for system and storage administrators who monitor and direct system confi

Figures1 Unidirectional 1-to-1 Remote Copy...162 Bidirectional 1-to

This relationship between target definitions and remote-copy system pairs is true for all validremote-copy configurations:• unidirectional• bidirectio

System Roles and Direction of Data FlowUnder normal operating conditions, the natural direction of data flow is from the primary volumegroup to the se

13 Remote-Copy LinksRemote-copy links are divided into two main types: sending links and receiving links.You create links during remote-copy setup by

Figure 32 Supported Sending-Link ConnectionsLinks per Remote-Copy PairWhen you set up remote-copy links between a remote-copy pair, you must create on

Figure 33 Sending Links for a Remote-Copy Pair—ExampleFor additional information, see the following:• “Setting Up the Primary System for Unidirectiona

14 Working with Virtual VolumesFor more informationCommandDescription“Creating Remote-Copy VirtualVolumes” (page 105)Review virtual volume requirement

modified. To create target volumes automatically, use the admitrcopyvv -createvvcommand. See:Automatically Creating Target Volumes on the Secondary Sy

2. Convert the standard virtual volume to an FPVV and assign it to a CPG by issuing the followingcommand:# setvv -snp_cpg <snp_CPG> <VV_name&

Renaming Virtual VolumesTo rename a virtual volume, issue the following command:# setvv -name <new_name> <VV_name>where:• <new_name>

When volumes draw additional space from the CPG, the volumesbecome larger and require additional storage space.To expand a volume’s storage space, the

53 Handling Failover...14454 Communication Fail

15 Working with Volume GroupsSee the following:For more information about:“Why Use Volume Groups” (page 110)What volume groups are and why remote copy

How Volume Groups WorkHP 3PAR Remote Copy functions as if the virtual volumes in a volume group are related andtherefore ensures that the data in the

Rules for Forming Volume GroupsThe group of virtual volumes in a remote-copy volume group:• Must be located on the same storage system.• Can be logica

Natural Direction of ReplicationThe natural direction of remote-copy replication is from the primary volume group to the secondaryvolume group (from t

Figure 35 Volume Mapping between Bidirectional Primary and Backup SystemsChanging the Remote-Copy Mode for a Volume GroupFor most remote-copy configur

where <pattern> represents the pattern of volume group names (for example, Group*).3. To start the primary volume group, issue either of the fol

Synchronous Mode• In synchronous mode, you must specify the -orvd option to manually force a fullsynchronization between volume groups in synchronous

Changing Secondary Volume Groups to PrimaryIf a primary system or primary volume fails, you can change secondary volume groups to primaryvolume groups

Setting Volume Group PoliciesFor more information:Policy Description“Automatically Restarting Volume Groups” (page 119)Automatically restart after cer

Automatically Restarting Volume GroupsBy default, remote-copy volume groups do not restart automatically.• To allow automatic restarts after all links

Part I Setting the Foundation for Remote Copy“Planning Your Remote-Copy Setup Strategy” (page 13)1. Plan your setup strategy.“Selecting a Remote-Copy

• To return the volume groups to the default policy and generate alerts for asynchronous periodicresynchronizations that do not complete within the sy

6. After the active cache update on the primary system is complete and the primary systemreceives the backup system’s acknowledgment, the primary arra

1. The host sends a write request to the primary system.2. As soon as the data is written into cache on the primary system, HP 3PAR Remote Copyacknowl

For additional information, see the following:• “Working with Volume Groups” (page 110)• “Volume Group Modes” (page 120)Types of SynchronizationRemote

3. When it is time to perform a resynchronization between the primary and secondary remote-copygroups, HP 3PAR Remote Copy creates snapshots as descri

4. Saves the active snapshot, which enables HP 3PAR Remote Copy to perform a fastresynchronization later (Figure 40 (page 126)).Figure 39 Full Synchro

Figure 40 Fast Resynchronization Using Resynchronization Snapshots6. Deletes the old resynchronization snapshot.Setting Up Resynchronization for Async

Considerations for Setting the Synchronization PeriodCAUTION: When you set the synchronization period, you must allow sufficient time for the volumegr

16 Optimizing Performance“Optimizing RCIP Throughput by Increasing MTU Size”(page 128)To optimize RCIP throughput, increase MTU size.“Avoiding RCIP Th

To optimize throughput, consider increasing the MTU size to 9000 bytes.“Checking RCIP Link Throughput and Latency” (page 33)1. Check link throughput.

1 Planning Your Remote-Copy Setup StrategySetup Instructions and ResourcesDepending on your comfort level and familiarity with HP 3PAR Remote Copy Sof

where:• <tput_value>—Maximum throughput limit and size unit. By default, throughput ismeasured in KBps. Optionally, you can specify the size uni

1. Set the local switch port to full duplex, if possible.To change the switch port settings, refer to your switch manufacturer’s documentation.2. To s

17 HP 3PAR Peer Persistence with Transparent FailoverHP 3PAR Peer Persistence is available for VMware configurations from version 3.1.2 MU2 of theHP 3

Table 7 HP 3PAR Peer Persistence Switchover and Failover RequirementsDescriptionRequirementFailoverRemote-copy groups are configuredin synchronous mod

Figure 41 Quorum Witness Update and Status MessagingHP 3PAR Peer PersistenceHP 3PAR Peer Persistence enables the transparent migration of applications

HP 3PAR Peer Persistence failover can occur when either the array or the site becomes unavailable,as shown in Figure 42 (page 134) and Figure 43 (page

Figure 44 Manual Peer Persistence SwitchoverPerforming a switchover moves an entire VMware data store from one array to another. Executinga switchover

Figure 45 HP 3PAR Peer Persistence—Working PrincipleTo summarize the switchover:• Synchronous remote-copy volumes are created on source and target arr

Figure 46 Transition StatesAs the secondary path becomes activated, the following actions occur (see Figure 47 (page 138)):1. The target remote-copy g

1. The source-array target-port group state is changed to standby, and any blocked I/O is returnedto the host with a sense error:NOT READY, LOGICAL UN

General Considerations• Distance between the systems.Systems in the same room can be connected through Gigabit Ethernet (GigE) switches,or through FC

Figure 49 Role Reversal BeginsTo complete role reversal, the following actions occur (see Figure 50 (page 140)):1. When the primary-reversed and secon

The transparent switchover is complete (see Figure 51 (page 141)):• The system is now fully reversed and ready for another switchover request.• The sw

in a peer-persistence configuration, additionally configured for automatic failover, to post updatesregarding their operational presence and health.St

Figure 52 VMware vMSC Certification—Uniform Host AccessUniform host access provides:• Protection against storage failure• Active load balancingHandlin

Figure 53 Handling FailoverFor more information about handling peer-persistence failover, see “Quorum Troubleshooting”(page 267).Link and Communicatio

Figure 54 Communication Failure—Array to ArrayCommunication Failure—Site 1 to Quorum WitnessIn the event of a communication failure between site 1 and

Communication Failure—Site 1 to Quorum Witness and Site 1 to Site 2In a dual-network communication failure between site 1 and the quorum witness, and

Figure 57 Communication Failure—Site 2 to Quorum WitnessCommunication Failure—Site 2 to Quorum Witness and Site 1 to Site 2In a dual-network communica

No Communication on the Part of Quorum Witness with Site 1 and Site 2In a dual-network communication failure resulting in no communication between the

Figure 60 No Communication among Quorum Witness, Site 1, and Site 2When all communication fails, service to the host must be manually re-enabled by tr

Remote Copy and Virtual DomainsHP 3PAR Remote Copy checks for the presence of HP 3PAR Virtual Domains Software (domains)on the backup system to verify

The response to hosts with personas that do not support ALUA (persona 6, for example) will be asfollows:Sense Key 0x02 Not ReadyASC 0x04

18 Using Tape for Initial Synchronization and BackupFor the first data synchronization between systems, it can be faster to back up primary volumegrou

where <group_name> represents the name of the volume group to start (Group1).or:# startrcopygroup -pat <pattern>where <pattern> repr

19 Managing VASA-connected Remote-Copy Groups andVolumesvSphere Storage application program interfaces (APIs) for Storage Awareness (VASA) is a set of

• -snp_cpg <cpg_name> <target_name>:<cpg_name>...—The local snapshot CPGand target snapshot CPG that will be used for volumes that a

where -removevv <VV_name> <group_name> automatically removes the correspondingsecondary volume of the specified volume in the specified vo

20 Error HandlingSee the following:For information about:“Synchronization States” (page 156)Volume synchronization states“Timeouts for Remote-Copy Fai

HP 3PAR Remote Copy uses low timeout values for synchronous volume groups to ensure thatremote copy stops replication before the host I/O times out.Ti

Restarting Synchronous Volume Groups after Link RecoveryAfter you recover at least one link between the systems and remote copy marks the link(s) as U

Failure of a Storage SystemIf one of the systems in a remote-copy pair fails, the other system:• detects the failure as a concurrent failure of both c

2 Selecting a Remote-Copy ConfigurationReview the remote-copy configurations and select theconfiguration that is best suited for your system.• “Unidir

Read ErrorsRead Errors during Initial SynchronizationDuring initial synchronization:• Synchronous volume groups—HP 3PAR Remote Copy reads the base vol

To resynchronize the virtual volumes:• Wait for the next scheduled synchronization. If it completes successfully, HP 3PAR RemoteCopy marks the virtual

Part IV Recovering from DisasterSee the following:To recover and restore this configuration after theprimary system goes offline:“Using Peer Persisten

21 Using Peer Persistence for Non-disruptive Failover in1-to-1 Remote Copy in Geocluster EnvironmentsHP 3PAR Peer Persistence software allows you to f

NOTE: With the path_management policy enabled, host persona 11 configures the targetport groups of secondary volumes as STANDBY. With HP 3PAR OS 3.1.2

Figure 61 Bidirectional 1-to-1 Remote-Copy Configuration—ExampleExample secondary systemExample primary systemSystem2System1System name1196System IDSy

System Information during Normal OperationPrimary System (System1)When all systems are up, the showrcopy command output on the primary system (System1

Name Target Status Role Mode Optionssync_group_1.r96 System1 Started Secondary Sync LocalVV ID Re

Switchover1. Before performing a switchover operation, confirm that the remote-copy group is started andsynced.2. Reverse the role of all volume group

3. Verify the reversal of volume group roles by performing the following actions:a. Issue the showrcopy command on the target system (System2).b. In t

Figure 2 Bidirectional 1-to-1 Remote CopyN-to-1 ConfigurationAn N-to-1 remote-copy configuration is composed of multiple remote-copy pairs. A maximum

• Confirm that the host persona is configured as persona 11 (VMware).• Confirm that the associated hosts are operating with VMware ESX release 5.0.170

22 Disaster Recovery for 1-to-1 ConfigurationsBefore you begin disaster recovery:“Bidirectional 1-to-1 Remote-Copy Configuration—Example” (page 172)1.

Figure 62 Bidirectional 1-to-1 Remote-Copy Configuration—ExampleExample backup systemExample primary systemSystem2System1System name1196System IDSyste

System Information during Normal OperationPrimary System (System1)When all systems are up, the showrcopy command output on the primary system (System1

Name Target Status Role Mode Optionssync_group_1.r96 System1 Started Secondary Sync LocalVV ID Re

Name Target Status Role Mode Optionssync_group_2 System1 Stopped Primary Sync LocalVV ID RemoteVV

2. Verify the reversal of volume group roles:a. Issue the showrcopy command on the failover system (System2).b. In the Group Information area, check t

a. Issue the showrcopy command on the failover system (System2).b. Verify the following:• The Status of the target (recovered) system is ready• The St

Link InformationTarget Node Address Status OptionsSystem2 0 10.100.33.11 Up System2 1 10.101.33.11 Up receive 0 1

6. Verify that failback is complete and the natural direction of data flow is restored:a. Issue the showrcopy command on the failover system (System2)

Figure 3 (page 18) illustrates an N-to-1 configuration where:• Bidirectional remote copy is maintained between remote-copy pair System3 and System4.Un

d. Verify that the Role of the primary volume group is Primary.The following showrcopy output for System1 shows that role for group sync_group_1is cor

23 Disaster Recovery for M-to-N ConfigurationsDisaster Recovery for N-to-1 ConfigurationsBefore you begin disaster recovery:“Unidirectional N-to-1 Rem

Example backup systemExample second primarysystemExample first primary systemVirtual volumes and volumegroups:• Two corresponding VVs:remotevv.0 andre

First Primary System (System2)When all systems are up, the showrcopy command output on one of the primary systems (System2)looks like the following ou

System Information during Failover for N-to-1 ConfigurationsSynchronous Mode for N-to-1 ConfigurationsWhen a failure occurs on systems with synchronou

Failover for N-to-1 Configurations1. Reverse the role of all volume groups on the system that is still in normal operation (the failoversystem). On th

2. Verify the reversal of volume group roles.a. Issue the showrcopy command on the failover system (System1).b. In the Group Information area, check t

2. When both systems in the remote-copy pair are ready to resume normal operation, reversethe natural direction of data flow and resynchronize the sys

remotevv.2 23 localvv.0 385 Synced Thu Dec 14 18:23:45 PST 2006 remotevv.3 24 localvv.1 386 Synced Thu Dec 14 18:23:4

4. On the recovered system (System2), issue the following command to verify that the volumegroups are started:# showrcopy groups <group_name>whe

Figure 4 1-to-N Remote CopyM-to-N ConfigurationIn an M-to-N remote-copy configuration, bidirectional data replication takes place in a 4x4 fan-in,fan-

6. Verify that failback is complete and the natural direction of data flow is restored:a. Issue the showrcopy command on the failover system (System1)

d. Verify that the Role of the primary volume groups is Primary.The following example of the showrcopy command output for System2 shows that theRole o

Figure 64 Unidirectional 1-to-N Remote-Copy ConfigurationExample second backupsystemExample first backup systemExample primary systemSystem3System2Sys

Target Node Address Status OptionsSystem2 0 10.100.33.61 Up System2 1 10.101.33.61 Up System3 0 10.100.33.11 Up

Status: Started, NormalTarget InformationName ID Type Status Options Policy System1 4 IP ready mirror_configLink InformationTarge

System1 1 10.101.33.96 Down receive 0 10.100.33.96 Down receive 1 10.101.33.96 Down Group InformationName Target St

2. Verify the reversal of volume group roles.a. Issue the showrcopy command on the failover system (System2).b. In the Group Information area, check t

b. Verify that for the target (recovered) system (System1):• the Status of the target system is ready• the Status of all sending links is Up• the Sync

System2 1 10.101.33.61 Up System3 0 10.100.33.11 Up System3 1 10.101.33.11 Up receive 0 10.100.33.61 Up r

5. Restore the natural direction of data flow to the remote-copy pair. On the failover system(System2), issue the following command:# setrcopygroup re

© Copyright 2014 Hewlett-Packard Development Company, L.P.Confidential computer software. Valid license from HP required for possession, use or copyin

Figure 5 M-to-N Remote-Copy Configuration—ExampleSynchronous Long Distance ConfigurationA synchronous long distance (SLD) remote-copy configuration is

6. Verify that failback is complete and the natural direction of data flow is restored:a. Issue the showrcopy command on the failover system (System2)

Name Target Status Role Mode Optionsperiodic_group_1 System2 Started Primary Periodic Last-Sync Thu De

24 Disaster Recovery for SLD ConfigurationsBefore you begin disaster recovery:“SLD Remote-Copy Configuration” (page 203)1. Review the example remote-c

Figure 65 SLD Remote-Copy ConfigurationExample asynchronousperiodic backup systemExample synchronousbackup systemExample primary systemSystemCSystemBS

SystemB 1:2:1 21210002AC0000C3 Up SystemC 0:3:1 10.100.33.63 Up SystemC 1:3:1 10.101.33.63 Up receive 0:2:1 2

Name Target Status Role Mode Options multi.1.r96 SystemC Backup Secondary Periodic Period 5m, over_per

Asynchronous Periodic Mode System Information during FailoverWhen a failure occurs on systems with asynchronous periodic mode volume groups such that

SystemA 42 IP failed mirror_configSystemB 44 IP ready mirror_configLink InformationTarget Node Address Status OptionsSyste

Figure 66 Failing Over to the Synchronous Backup SystemFollow these failover and recovery guidelines to recover from system failure.Failover in SLD Re

~/bin# waittask -v 5745 Id Type Name Status Phase Step -------StartTime------- ------FinishTime-------5745 remote_copy_failove

SLD configurations differ from 1-to-N configurations as follows:• In SLD configurations, remote-copy volumes on the primary system are in a remote-cop

3. Verify the reversal of volume group roles:a. Issue the showrcopy command on the failover system (SystemB).b. In the Group Information area, check t

4. Verify that the asynchronous periodic backup system (SystemC) is now the backup system.a. Issue the showrcopy command on the asynchronous periodic

where <target_name> represents the name of the target system with which to synchronize(SystemA).The setrcopygroup recover command:• Switches the

a. Issue the showrcopy command on the failover system (SystemB).b. Verify the following for the target system (SystemA):• The Status of the target sys

The following example shows the showrcopy command output for SystemA:# showrcopyRemote Copy System InformationStatus: Started, NormalTarget Informatio

Target Node Address Status OptionsSystemA 0 10.100.33.96 UpSystemA 1 10.101.33.96 UpSystemB 0 10.100.33.195 Up SystemB 1

CAUTION: The volume groups must be started before you continue to the next step and issuethe setrcopygroup restore command.5. To restore the natural d

6. Verify that the SLD setup is restored and operating normally.a. Issue the showrcopy command on the recovered system (SystemA) and verify that it ha

# showrcopyRemote Copy System InformationStatus: Started, NormalTarget InformationName ID Type Status Options Policy SystemA 46 FC

c. Issue the showrcopy command on the asynchronous periodic backup system (SystemC)and verify that it is in the role of backup system.The following ex

Because SLD configurations include a standby link, as illustrated in Figure 6 (page 21), thisconfiguration allows for multiple backup and recovery sce

Figure 67 Failing Over to the Asynchronous Periodic System with Data TransferFollow these failover and recovery guidelines to fail over to the asynchr

2. Verify that failover tasks complete successfully (failover of individual groups can still fail).For example, failing over to the asynchronous perio

3. Verify that the newer data is being pushed to the failover system.a. While remote copy progresses through the failover tasks, issue the showrcopy c

# showrcopyRemote Copy System InformationStatus: Started, NormalTarget InformationName ID Type Status Options Policy SystemA 42 IP failed

4. Verify the reversal of volume group roles.a. Issue the showrcopy command on the failover system (SystemC).b. In the Group Information area, check t

# showrcopyRemote Copy System InformationStatus: Started, NormalTarget InformationName ID Type Status Options Policy SystemA 46 FC

In the following example, HP 3PAR Remote Copy copies data from the virtual volumes involume group multi.1.r96 on SystemC to multi.1 on SystemA.• Start

b. Verify the following the target system (SystemA):• The Status of the target system is ready.• The Status of primary system’s sending links (SystemA

Remote Copy System InformationStatus: Started, NormalTarget InformationName ID Type Status Options Policy SystemB 58 FC ready 2FF

f. Verify the following:• The Status of the target system (SystemA) is ready.• The Status of the primary system’s sending links (SystemA) is Up.The fo

3 Selecting Network Connections1. Review possible network connections.• “IP Networks” (page 23)• “Fibre Channel Networks” (page 24)• “Fibre Channel ov

4. To verify that the volume groups are started, issue the following command on the recoveredsystem (SystemA):# showrcopy groups <group_name>whe

SystemA.2009-06-29 14:56:56 PDT Syncing Synchronization started for group multi.1.r96 to target SystemB.2009-06-29 14:56:56 PDT Waiting Waiti

6. Verify that the SLD setup is restored and operating normally.a. Issue the showrcopy command on the recovered system (SystemA) and verify that it ha

# showrcopyRemote Copy System InformationStatus: Started, NormalTarget InformationName ID Type Status Options Policy SystemA 46 FC

c. Issue the showrcopy command on the asynchronous periodic backup system (SystemC)and verify that it has resumed the role of backup system.The follow

Failing Over to the Asynchronous Periodic Backup System (No DataTransfer)If the primary system fails or is taken offline, you can failover the storage

Failover1. Reverse the role of all volume groups on the asynchronous periodic backup system (the failoversystem). On the failover system (SystemC), is

3. Verify the reversal of volume group roles:a. Issue the showrcopy command on the failover system (SystemC).b. In the Group Information area, check t

# showrcopyRemote Copy System InformationStatus: Started, NormalTarget InformationName ID Type Status Options Policy SystemA 46 FC

The setrcopygroup recover command:• Switches volume groups on the recovered system (SystemA) to secondary mode• Copies data from the reversed volume g

Fibre Channel NetworksRCFC can be set up only on storage systems that communicate over FC storage area networks (FCSAN).• Be sure you understand the F

multi.15915 rcpy.222.103.58 snp vcopy testvv.3 103 RO normal -- -- -- 1024 snapmulti.13. Verify that the primary system is up and the remote-copy pair

• The Role of the synchronizing volume groups is Secondary-Rev.• The SyncStatus of all volumes in the Secondary-Rev volume groups is Synced.The follow

f. Verify the following:• The Status of the target system (SystemA) is ready.• The Status of the primary system’s sending links (SystemA) is Up.The fo

4. To verify that the volume groups are started, issue the following command on the recoveredsystem (SystemA):# showrcopy groups <group_name>whe

2009-06-29 14:56:56 PDT Syncing Synchronization started for group multi.1.r96 to target SystemB.2009-06-29 14:56:56 PDT Waiting Waiting for sy

6. Verify that the SLD setup is restored and operating normally.a. Issue the showrcopy command on the recovered system (SystemA) and verify that it ha

# showrcopyRemote Copy System InformationStatus: Started, NormalTarget InformationName ID Type Status Options Policy SystemA 46 FC

c. Issue the showrcopy command on the asynchronous periodic backup system (SystemC)and verify that it has resumed the role of backup system.The follow

Figure 69 One Backup System Failure with SLD Remote CopyIn this example, when SystemB is restored, HP 3PAR Remote Copy transfers data from SystemA toS

Figure 70 (page 249) illustrates an SLD configuration in which the primary system, SystemA, goesdown. In this example, SystemB becomes the temporary p

4 Gathering Setup Information1. Contact your network administrator to obtain:• IP addresses or 64-bit World Wide Name (WWN) address information for th

In this example, when SystemA is restored, HP 3PAR Remote Copy performs the following actions:• Transfers data from SystemB to SystemA.• Restores Syst

Part V ReferenceSee the following:For information about:“Quick Setup Guide” (page 252)Quick setup steps for RCIP, RCFC, and remote-copyconfigurations“

25 Quick Setup GuideSee the following:To set up:“Quick Setup: RCIP Transport Layer” (page 253)Transport layer for RCIP“Quick Setup: RCFC Transport Lay

Quick Setup: RCIP Transport LayerTable 8 Quick Setup Steps for RCIPBackup SystemPrimary SystemReferenceStepcontrolport rcip addr[-f] <port_IP>&l

Quick Setup: RCFC Transport LayerTable 9 Quick Setup Steps for RCFCBackup SystemPrimary SystemReferenceStepcontrolport offline<N:S:P>controlport

Quick Setup: Unidirectional Remote CopyTable 10 Quick Setup Steps for Unidirectional Pairs: 1-to-1, N-to-1, and 1-to-N ConfigurationsBackup SystemPrim

Table 10 Quick Setup Steps for Unidirectional Pairs: 1-to-1, N-to-1, and 1-to-N Configurations(continued)Backup SystemPrimary SystemReferenceStep• If

Quick Setup: Synchronous Long Distance Remote CopyTable 12 Quick Setup Steps for Synchronous Long Distance ConfigurationsPeriodic SystemSync SystemPri

Table 12 Quick Setup Steps for Synchronous Long Distance Configurations (continued)Periodic SystemSync SystemPrimary SystemReferenceStep<periodic_t

26 Quick Disaster Recovery GuideSee:To recover from primary system downtime for:“Quick Recovery: 1-to-1 and M-to-N Remote-CopyConfigurations” (page 25

5 Setting the Transport Layer“Transport Layer Considerations” (page 26)Review transport layer requirements.Set the transport layer.• “Setting Up Remot

Quick Recovery: Synchronous Long Distance ConfigurationsTable 14 Quick Recovery Steps for Synchronous Long Distance ConfigurationsBackup SystemFailed/

27 Using Remote Copy CommandsFor information about the following commands, see the HP 3PAR Command Line Interface Reference.To access this document, g

CommandDescriptionsrstatrcvvView historical performance data reports for remote-copyvolumesstartrcopyEnable remote copy on a storage systemstartrcopyg

28 HP 3PAR Quorum Witness DeploymentThis appendix describes how to deploy HP 3PAR Quorum Witness for peer-persistence failover.See “HP 3PAR Peer Persi

4. Type your new password, and then retype the password to confirm.5. Select Configure the Device, then select the device to configure, either eth0 or

nameserver 192.168.1.36. Restart the network service by issuing the following command:service network restart7. Verify IP connectivity to arrays.Quoru

1. Check the status of the quorum witness by issuing the showrcopy -qw command, and ensurethat the quorum-witness state has advanced to the not-starte

To remove a quorum witness, you must first stop it; see “Stopping a Quorum Witness” (page 266).Then issue the remove command:# setrcopytarget witness

Table 15 Quorum Status Qualifiers (continued)DescriptionQuorum Status Qualifieris either in the Initializing state, or is not sufficientlyfar enough i

Table 15 Quorum Status Qualifiers (continued)DescriptionQuorum Status QualifierA locally detected quorum-announcer status inconsistencywill typically

Storage and HP 3PAR 10000 StoreServ Storage systems, individual ports can be configured forRCFC.Remote-copy links in asynchronous periodic mode do not

29 Remote-Copy SnapshotsSee the following:For information about:“Why Use Snapshots” (page 270)What snapshots are and why remote copy uses them“Volume

For additional information, see the following:• “Common Provisioning Groups” (page 108)• “Converting Standard Virtual Volumes” (page 106)Snapshots in

Snapshots Taken before or during Disaster RecoveryIf the Backup System FailsIf the backup system fails, or all communication links to the backup syste

• When the resynchronization completes, remote copy deletes the old snapshots on the primaryand the backup systems.• After resynchronization, the base

30 Support and Other ResourcesContacting HPFor worldwide technical support information, see the HP support website:http://www.hp.com/supportBefore con

See:For information about:HP 3PAR-to-3PAR Storage Peer Motion GuideMigrating data from one HP 3PAR storage system to anotherHP 3PAR Secure Service Cus

See:For information about:Planning for HP 3PAR storage system setupHardware specifications, installation considerations, power requirements, networkin

Typographic conventionsTable 16 Document conventionsElementConventionBold text• Keys that you press• Text you typed into a GUI element, such as a text

31 Documentation feedbackHP is committed to providing documentation that meets your needs. To help us improve thedocumentation, send any errors, sugge

GlossaryAasynchronousperiodic modeData replication in which the primary and secondary volumes are resynchronized at set times—forexample, when schedul

Figure 10 Network and RCIP Setup—Example Initial-Pair WorksheetRCIP Additional Remote-Copy Pairs WorksheetFigure 11 (page 29) is a worksheet for setti

synchronous mode Data replication that writes data to the primary and secondary sites simultaneously over a network,such that data remains current bet

IndexSymbols1-to-1quick recovery, 259remote copy, 161-to-1 remote copyillustration, 45setupbidirectional, 40unidirectional, 401-to-Nquick recovery, 25

dismissrcopylink, 30, 51, 66, 81, 261dismissrcopytarget, 261dismissrcopyvv, 53, 68, 83, 90, 261-removevv, 154growvv, 105, 107, 155, 261histrcvv, 95, 2

consistency groupsconfiguring the synchronization period, 126loss of sync between, 157synchronization of new consistency groups, 123controller node po

FPVVcreating, 151limits, 106new or empty, 152free snapshot data space, 108fully-provisioned virtual volume see FPVVGgatewayIP address, 25RCIP, 31Gigab

MTU, 32increasing, 33NN-to-1quick recovery, 259N-to-1 remote copyconfiguration, 17restrictions, 43natural direction of data flow, 101networkbandwidth,

bidirectional 1-to-1, 16configurations, 16N-to-1, 17over Fibre Channel, 24setup, 254pairs, 16, 99removing, 94setup, 13setup strategy, 13SLD, 20SLD res

176, 177, 179, 182, 183, 184, 186, 187, 188, 189, 190, 191, 192, 193, 194, 196, 197, 198, 200, 203, 204, 205, 206, 210, 211, 213, 214, 215, 217, 219,

Ttapefor synchronization and backup, 151not supported with SLD, 151targetbidirectional remote-copy pair, 100definitions, 99target definition, 43target

virtual machinemigration, 136Red Hat Linux, 133virtual volumes, 105creating, 105limits, 106limits in volume groups, 111primary volume, 113remote-copy

Figure 11 Network and RCIP Setup Information—Example Additional Pairs WorksheetSetting Up Remote Copy over IP 29

ContentsI Setting the Foundation for Remote Copy...121 Planning Your Remote-Copy Setup Strategy...

Configuring Ports for RCIPFor RCIP configurations, each link between a remote-copy pair is a logical link between a controllernode on one storage syst

4. Repeat Step 1 through Step 3 for each Ethernet port on each HP 3PAR StoreServ Storage usedin the remote-copy configuration.5. Issue the following c

Verifying That the Servers Are ConnectedPerform the following steps to verify that the servers are connected:1. Issue the following command:# controlp

Checking RCIP Link Throughput and LatencyBy default, the cli checkrclink command uses port 5001. There is, however, an optionalparameter that allows y

1. Issue the controlport rcip mtu command:# controlport rcip mtu 9000 <N:S:P>where <N:S:P> represents the location of the GigE port, expre

Figure 12 Network and RCFC SetupTo configure an RCFC transport layer, use the worksheet in Figure 13 (page 35)Figure 13 Network and RCFC Setup Informa

To configure the ports, use the controlport rcfc command.CAUTION: Each pair of RCFC links must exist in an exclusive zone. Fabric zones cannot beshare

2. Issue the following command:# controlport config rcfc -ct point -f <N:S:P>where:• -ct point—point-to-point mode.• <N:S:P>—FC adaptor po

Checking RCFC Link Throughput and LatencyCAUTION: Do not check the links if CPU usage is already close to 100%. The link checktemporarily increases CP

Part II Setting Up the Remote-Copy ServersSee the following:To set up:“Setting Up N-to-1 Remote Copy” (page 40)M-to-N remote copy• 1-to-1 remote copy•

1-to-N Remote-Copy Considerations...44M-to-N Remote-Copy Examples...

6 Setting Up M-to-N Remote Copy1-to-1 Remote CopyFollow these guidelines for setting up 1–to-1 remote copy.To set up 1-to-1 remote copy in a unidirect

Table 2 Setting up N-to-1 Remote CopyFor information, see...Step“Requirements” (page 13)1. Verify that your setup plan meets the N-to-1requirements.“N

Table 3 1-to-N Remote CopyFor information, see...Step“Requirements” (page 13)1. Verify that your setup plan meets the 1-to-Nrequirements.“1-to-N Remot

synchronous/asynchronous periodic 1-to-1 remote-copy configuration is 2400 volumes in all(not 2400 volumes for synchronous mode and 6000 volumes for a

Figure 16 N-to-1 Remote-Copy ConfigurationIn an N-to-1 remote-copy configuration:• All of the remote-copy pairs can be bidirectional.• You can use RCI

Figure 17 1-to-N Remote-Copy ConfigurationIn a 1-to-N remote-copy configuration:• Primary systems with only two controller nodes must be one of the fo

Figure 18 Unidirectional 1-to-1 Remote-Copy Pair—ExampleExample backup systemExample primary systemSystem2System1System name125112System IDfour (only

Figure 19 N-to-1 Remote-Copy Configuration—ExampleExample backup systemExample primary systemSystem2System1System name125112System IDfour (only two ar

Figure 20 Example 1-to-N Remote-Copy ConfigurationOne of the remote-copy pairs can be bidirectional, and the other can be unidirectional. Bothremote-c

• <N:S:P>—Location on the current system (System1) that contains the IP connections tothe target system, expressed as node:slot:port• <link_I

11 Viewing Remote-Copy System Information...95How Remote Copy Tracks Synchronization Details...

where:• <target_name>—Name of the target system (System1)• FC—Defines link as an FC link• <node_WWN>—WWN of the node on the target system•

3. Repeat Step 1 and Step 2 on the primary system (System1).SolutionProblemYou encounter problemswith the links.1. Use the dismissrcopylink command to

1. Issue the following command to set up a volume group on the primary storage system (System1):# creatercopygroup <group_name> <target_name&

3. On the primary system (System1), add the pre-existing virtual volumes to the newly createdvolume group by issuing either of the following commands:

Selecting an Initial Replication Method for M-to-N Remote CopySelect the method to use to start replicating the data in the primary volume groups:• Di

Name Target Status Role Mode Optionssync_group_1 System2 Started Primary Sync LocalVV ID RemoteVV

7 Setting Up SLD Remote Copy“Requirements” (page 13)1. Verify that your setup plan meets the SLD requirements.“SLD Remote-Copy Considerations” (page 5

• With HP 3PAR OS 3.1.3 and later, you can create four target systems per source system.• If you are using two-node systems for your SLD configuration

◦ <node_WWN>—WWN of the target node (node 2FF70002AC0000C3 on SystemB).◦ <N:S:P>:<WWN>—FC adapter port on the primary system (System

3. To define the targets for the asynchronous periodic backup system (SystemC) on the primarysystem (SystemA), issue the following command on the prim

Setting Volume Group Policies...118Automatically Restarting Vo

2. Define a target (for example, the primary system SystemA) on the backup system. On thebackup system (SystemB), issue the following command:• For RC

where:• <target_name>—Name of the remote-copy target system (SystemC).• IP—Specifies an IP link.• <N:S:P>:<IP_addr>—Node on the back

6. Define a second target system (for example, the synchronous backup system SystemB) for thecurrent backup system. On the current backup system (Syst

Figure 21 SLD Remote-Copy Configuration—ExampleExample backup system 2Example backup system 1Example primary systemSystemCSystemBSystemASystem namefou

The primary system and backup system 1 (SystemA and SystemB) are located near each other.• SystemA and SystemB are connected through RCFC or RCIP.• Re

2. Verify the following:• The Target Information area shows a Status of ready.• The Link Information area shows a Status of Up for all the links.The f

4. Repeat Step 1 and Step 2 on the other backup system (SystemC).The following example shows the showrcopy command output for SystemC:Remote Copy Syst

1. Create a volume group on the primary system (SystemA):# creatercopygroup <group_name> <target_name>:sync <target_name>:periodicwh

2. On the primary system (SystemA), add the pre-existing virtual volumes to the newly createdvolume group. Issue either of the following commands:# ad

Starting Initial Replication for SLD Remote Copy: Copying Data Directlyfrom Primary Volume GroupsFollow these steps to start initial replication:1. To

Setting Associated Local and Remote CPGs for Remote-Copy Groups...153Automatically Creating Target Volumes on the S

3. On all three systems, verify that the SyncStatus column in the Group Information areadisplays Synced.The following example shows the showrcopy grou

Failover and Failback Behavior of Individual Remote-Copy Groups in SLDConfigurationIndividual remote-copy groups can be failed over without forcing a

Figure 23 Failure of an Individual Volume Group in SLD Remote-Copy ConfigurationSystemA fails over to SystemB. SystemB becomes the primary-reverse sys

between SystemB and SystemA, and in asynchronous periodic mode from SystemB to SystemC,as shown in Figure 25 (page 73).Figure 25 SystemA Becomes the S

Figure 26 SystemA-to-SystemC FailoverAfter all updates have been sent to SystemC, SystemB then reverts to being the secondary system.After SystemC rec

When the setrcopygroup recover command is issued, SystemA comes back online and isnow the secondary-reverse system, receiving updates from SystemC in

8 Unidirectional ConfigurationFollow these guidelines to set up a unidirectional configuration:1. Verify that your setup plan meets remote-copy requir

Figure 29 Unidirectional Remote-Copy Configuration—ExampleExample backup systemExample primary systemSystem2System1System name125112System IDfour (onl

• For RCFC remote-copy pairs:# showrctransport -rcfcFor more information, see “Verifying That the Servers Are Connected” (page 32).Setting Up the Prim

2. To define a target on the primary storage system (System1), issue the following command:• For RCIP:# creatercopytarget <target_name> IP <N

Second Primary System (System3)...183System Information during Failover for

2. To define a target on the backup system (System2), issue the following command:• For RCIP:# creatercopytarget <target_name> IP <N:S:P>:

The following example shows the showrcopy command output for System2:Remote Copy System InformationStatus: Started, NormalTarget InformationName ID

Creating Volume Groups for Unidirectional Remote CopyRemote-copy volume groups are pairs of virtual volume sets that are logically related and that ar

where:• <period_value>s|m|h|d—Synchronization time period (time between start ofresynchronizations), followed by time unit: s (seconds), m (minu

Selecting an Initial Replication Method for Unidirectional Remote CopySelect the method for replicating the data in the primary volume groups:• Direct

4. Repeat the steps in “Starting Initial Replication for Unidirectional Remote Copy: Copying DataDirectly from Primary Volume Groups” (page 84) for al

9 Bidirectional ConfigurationSee:1. Verify that your setup plan meets the remote-copyrequirements.• “Requirements” (page 13)• “N-to-1 Remote-Copy Cons

Figure 30 Bidirectional Remote-Copy Configuration—ExampleExample backup systemExample primary systemSystem1System2System name112125System IDSystem2Sys

Setting Up the Systems for Bidirectional Remote CopyIn bidirectional remote copy, the backup system (System2) also contains primary volume groupsand t

3. (Optional). If you are creating a volume group that uses periodic mode, you can set upautomatic synchronization.CAUTION: Do not set a resynchroniza

27 Using Remote Copy Commands...261About the Remote-Copy Commands...

4. On the primary system (System2), add the pre-existing virtual volume(s) to the newly createdvolume group (Group2) by issuing either of the followin

• Through tape. Using tape for the initial backup can be faster if you are copying very largevolumes.“Using Tape for Initial Synchronization andBackup

Part III Managing Remote CopySee the following:For information about:“Viewing Remote Copy System Information” (page 95)Viewing configuration and synch

10 Stopping, Starting, or Removing Remote CopyFor more informationCommandManual Operation“Stopping and Restarting Remote Copy”(page 94)stoprcopygroups

Stopping and Restarting Remote CopyTo stop remote copy temporarily, issue the following command:# stoprcopy -stopgroupsTo restart remote copy, issue t

11 Viewing Remote-Copy System InformationFor more information:Command:Description“About the Remote-Copy Commands”(page 261)showrcopyConfiguration deta

To verify that a resynchronization took place, check the tasks by using the showtask command.For information about using the showtask command, see “Vi

• Step—For active remote-copy synchronization tasks, indicates the completed and total volumesize in MB, in the following format:<completed_size>

Task Management Commands for Remote-Copy SynchronizationsTable 4 Task Management CommandsDomain: Privilege LevelDescriptionCommandall: Super, Edit, Se

12 Remote-Copy Pairs and TargetsSee the following:For more information about:“Remote-Copy Pairs” (page 99)The foundational remote-copy system relation