If disk and, or system failures occur during the disk group split operation, Storage Foundation generally performs any necessary recovery operations automatically. The recovery procedure performs any operations necessary to restore the disk group to a consistent state. On successful recovery, the disk groups are again available for use.
For example, if there is a power outage during the split operation, it results in incomplete information on the disk and in the disk group. The recovery mechanism ensures that when the host boots up the next time or if the disk group is imported on a host (the same host or another host), the inconsistencies in the information are removed. Depending on the amount of processing completed by the split operation at the time of the failure, the recovered disk group is either in the state it was in before the split operation or be successfully split into two disk groups.
Recovery has the following two modes:
Automatic recovery
If it is possible to restart the system, Storage Foundation attempts automatic recovery after the restart. If the automatic recovery process is able to detect sufficient information to perform the recovery, the recovery is successful and the disk group appears in a normal state after the automatic recovery process.
Manual recovery
In the following situations, it is necessary to do manual recovery:
If the automatic recovery runs after the restart and there is not sufficient information for recovery, the disk group is in a Deported state and displays an alert icon (a yellow caution icon).
If you cannot restart the system because it has failed completely, then you must move the disk group's disks to another system and then do a manual recovery. In this situation, it is recommended that you move the entire source disk group to another computer, rescan, and implement the recovery manually.
It becomes clear that a disk group needs manual recovery after a split command when Storage Foundation does not allow imports of the disk group to succeed or it blocks the join of a disk group. You receive an error message indicating that the disk group needs recovery.
Recovery should be done as soon as possible to preserve the integrity of the data.
You can perform manual recovery through the GUI as described in the following procedure.
You can also perform manual recovery through the command line.
On the successful completion of the manual recovery operation, you can perform all Storage Foundation operations. The only exception is that if a host crashes during a particular stage of the split operation, one or more of the disk group's disks and volumes may come up in an unhealthy state after recovery. In that situation, use the Reactivate Disk command on the unhealthy disks and then the Reactivate Volume command on each of the unhealthy volumes that are involved in the split command. After running these commands, the disk group's disks and volumes should be healthy.