Automated read-only recovery

Automated read-only (ROR) recovery is the process by which hierarchical storage management (HSM) recalls all active data from a particular physical volume that has exceeded its error thresholds, encountered a permanent error, or is damaged.

Some HSM functions such as reclaim or secure data erase will designate a physical volume as ROR for recovery. HSM requires that a certain number of drives be available for a period of time before ROR is allowed to proceed. At least one drive must be free for five minutes, or two drives must be free for any period of time. If either of these criteria is not met, HSM will defer the ROR recovery of that physical volume to the next hourly cycle.

In a TS7700 Grid configuration, if a virtual volume cannot be recovered from the ROR volume and a valid copy exists on another TS7700 Cluster, the remote virtual volume will be copied to the cluster performing the recovery and afterwards migrated to another physical volume. In this way, the previously unrecoverable copy is recovered. If the virtual volume cannot be recovered, an intervention is reported.