During data copy, if a failure occurs in a storage node other than the initiator node, the failure will result in reduced redundancy of the existing data.
To handle the failure, perform the following.
-
Verify whether data copy is running.
For details about how to verify whether data copy is running, see step 3 of Stopping data migration.
If data copy is running, go to step 2. If data copy is canceled, perform the procedure of Prioritizing redundancy recovery.
-
Perform one of the following:
-
Prioritize recovery of redundancy by canceling data copy and then recover the faulty storage node. (Recommended)
For the detailed procedure, see Prioritizing redundancy recovery.
-
Prioritize completion of data copy, and then take action for the storage node failure.
For the detailed procedure, see Prioritizing completion of data copy.
-
If the storage node auto-recovery function is enabled, the node is automatically recovered. Therefore, ongoing data copy is automatically canceled.
For the detailed procedure after auto recovery, see step 4 and its subsequent steps in Prioritizing redundancy recovery.
CAUTION:If completion of data copy is prioritized, redundancy of the existing data is reduced until the system recovers from the storage node failure. Prioritize redundancy recovery because the longer the redundancy-reduced state continues, the higher the operation shutdown risk becomes.
In the following cases, prioritizing completion of data copy causes no problems:
-
No storage node other than the initiator node has volumes (e.g., immediately after installation).
-
Although storage nodes other than the initiator node have volumes, no server references those volumes.
-
Immediately before completion of data copy
-