How the RECOVER utility performs fallback recovery

If neither image copy is usable, RECOVER attempts to fall back to a previous recovery point. If the previous recovery point is a full image copy, the RECOVER utility uses the full image copy, any incremental image copies, and the log to recover. If a previous REORG LOG YES or LOAD REPLACE LOG YES was done, RECOVER attempts to recover from the log and applies any changes that occurred between the two image copies. If good full image copies are not available, and no previous REORG LOG YES or LOAD REPLACE LOG YES jobs were run, the RECOVER utility terminates. The RECOVER utility will not fall back to a system-level backup.

If one of the following actions occurs, the index remains untouched, and utility processing terminates with return code 8:

• RECOVER processes an index for which no full copy exists.
• The copy cannot be used because of utility activity that occurred on the index or on its underlying table space,

If you always make multiple image copies, RECOVER should seldom fall back to an earlier point. Instead, RECOVER relies on the backup copy data set if the primary copy data set is unusable.

In a JES3 environment, you can do a fallback recovery by issuing a JES3 cancel,s command at the time the allocation mount message is issued. This action might be necessary if a volume is not available or if you do not want the given volume.

RECOVER does not perform parallel processing for objects that are in backup or fallback recovery. Instead, the utility performs nonparallel image copy allocation processing of the objects. RECOVER defers the processing of objects that require backup or fallback processing until all other objects are recovered, at which time the utility processes the objects one at a time.