Figure 85 Package Failover upon entire DC1 failure

NOTE: In Figure 85 (page 466), the writes to the disk at DC2 are not accepted till the Delta Resync pair is re-synchronized when using 3DC CAJ/CAJ replication. So, in the event of 3DC DR software failing to resync the Delta Resync pair, the application package fails to come up. In this case, though the DC1-DC2 device group pair is in SSWS state at DC2, the writes to the disk are rejected. You must fix the issue and manually resync the Delta Resync pair using the command pairresync -g <device group name> –swapsbefore restarting the package.

Recovery at the third data center (DC3)

Recovery at the third data center (DC3) is automated, and is manually triggered using the Continentalclusters command cmrecovercl. For more information on using this command, see section Section (page 94).

The Three Data Center software issues the horctakeover command at DC3. This command results in a swap takeover only if the there is continues access synchronous replication is configured between DC1 and DC2 and the status of this sync pair at the site in the primary cluster, where the journal device group is configured, is either SVOL-SSWS or PVOL-PSUS. In all other situations, the horctakeover command results in an SVOL takeover.

NOTE: HP StorageWorks XP 3DC CAJ/CAJ replication requires that the Active-CAJ pair to be deleted to allow an application to use the data copy at DC3. As part of recovery, cmrecovercl issues the pairsplit –S command to delete the Acive-CAJ pair on both ends of the replication (i.e. at DC3 and either at DC2 or DC1). If it fails, then cmrecovercl issues the pairsplit –R command to delete the Acive-CAJ pair only on the DC3 side.

Performing Recovery Group Rehearsal in the 3DC DR Solution

The Continentalclusters feature DR Rehearsal is supported in the 3DC DR Solution except when using the Site Aware Disaster Tolerant Architecture (SADTA).

To configure and perform disaster recovery rehearsal in a 3DC environment, see Disaster Recovery Rehearsal in Continentalclusters white paper available at http://www.hp.com/go/ hpux-serviceguard-docs —> HP Serviceguard Continentalclusters.

Failback Scenarios

Failback from Data Center 3 (DC3)

After restoring the primary cluster from the disaster, you can move the application packages back to the primary cluster. Complete the following steps to move the application package back to DC1:

466 Designing a Three Data Center Disaster Recovery Solution