If you experience an outage that affects both Fabric Management Platforms (FMP), make sure the FMP that you restart first is the FMP that was running the Fabric Manager workload before the outage. Use the following procedure to correctly recover the Fabric Manager database.
Note: The following procedure assumes you do not have a backup.
Power on the FMP that was previously the primary FMP.
Make sure the cluster service is running:
# rcopenais status
If it is not running, restart it:
# rcopenais start
Once you can access the Fabric Manager user interface, make sure all platforms and partitions are visible.
Repeat steps 1 and 2 on the other FMP.
In some situations you may not be able to follow this sequence of restarting the FMPs. For example, the cluster service may be set to start at boot time and the FMP that was originally the primary FMP was not started first. In this situation, once both FMPs are booted and the cluster service is running, restore the Fabric Manager database from the master FMP node.