HP XP Racks - Manual (Page 65)

Journal pair after you resynchronize the ESAM P-VOL that accepts host I/O or after you
resynchronize the pair in the opposite direction.
Prerequisites
Make sure you have analyzed the failure messages to determine if all of the conditions required
to use resynchronization exist.
For more information, see “Determining which resynchronization recovery procedure to use”
(page 62) .
Procedure
1. Using multipath software, vary the owner path offline.
2. Swap and suspend the ESAM pair using the pairsplit -RS command.
3. Swap and suspend the Continuous Access Journal pair using the pairsplit -RS command.
4. Resynchronize the Continuous Access Journal pair in the opposite direction using the pairresync
-swaps command. The system copies data in the Continuous Access Journal S-VOL to the
ESAM S-VOL.
5. When the Continuous Access Journal pair is in PAIR status, swap and suspend the pair using
the pairsplit -RS command.
6. Resynchronize the Continuous Access Journal pair in the opposite direction using the pairresync
-swaps command.
7. When the Continuous Access Journal pair is in PAIR status, split the pair using the pairsplit -r
operation.
8. Resynchronize the ESAM pair in the opposite direction using the pairresync -swaps command.
The system copies data in the ESAM S-VOL to the ESAM P-VOL.
9. Using multipath software, vary the owner path online.
10. Using multipath software, vary the non-owner path offline.
11. When the ESAM pair is in PAIR status, swap and suspend the pair using the pairsplit -RS
command.
12. Resynchronize the ESAM pair in the opposite direction using the pairresync -swaps command.
13. Using multipath software, vary the non-owner path online.
14. When the ESAM pair is in PAIR status, resynchronize the Continuous Access Journal pair using
the Pairresync operation.
Postrequisites
After the ESAM pair finishes resynchronizing, you must:
•
Resynchronize the Continuous Access Journal pair.
•
Run RAID Manager commands to complete the task.
Recovering from path failures
You can recover an owner path, non-owner path, or both paths from failure by restoring the paths.
CAUTION: If failures occur in both the owner and non-owner paths, you must restore the paths
in the correct order to prevent unexpected failover or failback. Make sure you restore the path to
the storage system that resumes host operations before you restore the other path.
Recovering from path failures 65