Operating System Software Owner manual

Failover
1. Reverse the role of all volume groups on the asynchronous periodic backup system (the failover
system). On the failover system (for example, SystemC), enter:
# setrcopygroup failover -f -discard -t <target_name>
• <target_name> - Name of the target system that failed (for example, SystemA)
The setrcopygroup failover -f -discard -t command:
• Bypasses the data transfer step between the backup systems (SystemB and SystemC)
• Changes all secondary volume groups on the failover system (multi.1.r96 on SystemC)
to primary volume groups, and therefore reverses the direction of data replication
• Enables I/O to the reversed volume groups (multi.1.r96 on SystemC)
• Starts all volume groups on the synchronous backup system (SystemB) (if that backup
system is still up)
• Makes the asynchronous periodic backup system (SystemC) the temporary primary storage
system
• Makes the synchronous backup system (SystemB) the temporary backup system
• Synchronizes the synchronous backup system to the failover system (synchronizes SystemB
so that its data matches that on SystemC), potentially discarding recent data
2. Verify that failover tasks complete successfully (failover of individual groups can still fail).
For example, failing over to the asynchronous periodic backup system (for example, SystemC)
results in a task list that looks like:
# setrcopygroup failover -f -t -discard SystemA
failover started with tasks: 6740
root@xnodeeca5:~/bin# waittask -v 6740
Id Type Name Status Phase Step -------StartTime-------
------FinishTime-------
6740 remote_copy_failover multi.1.r96 done --- --- 2009-06-29 14:43:44 PDT
2009-06-29 14:45:03 PDT
Detailed status:
2009-06-29 14:43:44 PDT Created task.
2009-06-29 14:43:44 PDT Begin Failover operation on group multi.1.r96 started.
2009-06-29 14:43:47 PDT Promote Not all volumes of group multi.1.r96 were synced to the
same time point with target SystemA, snap volumes will be promoted to bring the group to
the last group-consistent state.
2009-06-29 14:44:05 PDT Switched Group multi.1.r96 was switched from secondary to primary.
2009-06-29 14:44:05 PDT Waiting Wait for target SystemB of group multi.1.r96 to complete
snapshot promotions.
2009-06-29 14:44:05 PDT Starting Starting group multi.1.r96 to target SystemB.
2009-06-29 14:44:09 PDT Started Remote copy was started for group multi.1.r96 and target
SystemB.
2009-06-29 14:44:09 PDT Completed Group multi.1.r96 failover completed.
168 Disaster Recovery for Synchronous Long Distance