HP StorageWorks XP Remote Web Console Software - Manual (Page 222)

For detailed information about status of data volumes, see “Pair Status” (page 36) .
Resuming Normal Operations at the Primary Site
Once the Universal Replicator for Mainframe volume pairs have been established in the reverse
direction, you are ready to resume normal operations at the primary site. The following procedure
explains how to resume normal operations at the primary site by using Business Continuity Manager.
Remember that the Universal Replicator for Mainframe terminology is now reversed: the original
primary data volumes are now secondary data volumes, and the original secondary data volumes
are now primary data volumes.
1. At the primary and secondary sites, make sure that all Universal Replicator for Mainframe
components are fully operational and are free from failures.
2. Make sure that the pair status of the primary and secondary data volumes in all Universal
Replicator for Mainframe pairs is Duplex. This indicates that the Universal Replicator for
Mainframe initial copy operations are complete and consistency is maintained.
3. Stop the applications at the secondary site.
4. Issue a request for splitting pairs to master journal groups (which were originally restore journal
groups); use Business Continuity Manager to execute the YKSUSPND FLUSH SVOL PERMIT
option on the master journal group (which was originally the restore journal group); YKSUSPND
is a command for splitting pairs. If an error occurs when splitting pairs, remove the error cause
and go back to step 1 after resuming your business task at the secondary site.
5. If no error occurs in step 4, wait until the suspension finishes. After the suspension finishes,
check whether there is a secondary data volume (which was originally a primary data volume)
whose pair status is other than Suspend (equivalent to SUSPOP with Business Continuity
Manager). If such a pair exists, remove the error cause and go back to step 1 after resuming
your business task at the secondary site.
6. If there is no secondary data volume (which was originally a primary data volume) whose
pair status is other than Suspend (equivalent to SUSPOP with Business Continuity Manager),
data in the primary data volumes are the same as data in the secondary data volumes, and
the secondary data volume (which were originally the primary data volumes) are usable.
Resume applications at the primary site.
7. Execute the YKSUSPND REVERSE command on the restore journal group (which was originally
the master journal group); YKSUSPND is an Business Continuity Manager command and
REVERSE is an option. Wait until the suspension completes.
8. After the suspension completes, execute the Business Continuity Manager YKRESYNC REVERSE
command on the restore journal group (which was originally the master journal group). This
reverses the primary data volumes and secondary data volumes to resynchronize the pairs
and restores the copy direction to its original direction.
For detailed information about how to use Business Continuity Manager, see the HP XP for Business
Continuity Manager Software User's Guide.
For detailed information about status of data volumes, see “Pair Status” (page 36) .
Disaster Recovery for Multiple Primary and Secondary Storage Systems
Consistency of the Data Update Sequence When a Disaster Occurs
When a failure occurs in some primary or secondary storage system, all journal groups in the
extended consistency group will be suspended and remote copy operations will be suspended.
The consistency of the data update sequence will be maintained if a failure occurs in a primary
storage system, but the consistency will not be maintained if a failure occurs in a secondary storage
system.
For example, in the case described in the following figure, if a failure occurs in primary storage
system 1 at 10:00 a.m., journal groups 1 to 3 will be suspended due to the failure, and the data
written from the host before 10:00 a.m. will be restored to the secondary data volumes in the
222 Usage Scenarios