Point-in-time recovery of the catalog, directory, and all user objects
Full recovery of the catalog and directory table spaces and indexes is strongly recommended. However, in some situations, you might need to do a point-in-time recovery. In this case, you should understand the implications and plan for this type of recovery.
When you recover the Db2 catalog, directory, and all user objects, consider the entire catalog and directory, including all table spaces and index spaces, to be one logical unit. Recover all objects in the catalog, directory, and all user objects to the same point of consistency. If you plan to do a point-in-time recovery of the catalog, directory, and all user objects, a separate quiesce of the DSNDB06.SYSTSCPY table space is required after a quiesce of the other catalog and directory table spaces.
- A complete referential integrity (RI) set. If the complete RI set is not recovered together, CHKP is not set on the dependents.
- A complete base and LOB set. If base and LOB objects are not recovered together, ACHKP or CHKP is not set.
If you perform a point-in-time recovery on catalog and directory table spaces, the indexes are placed in RBDP (rebuild-pending) status. Use the CHECK INDEX utility to determine whether an index is inconsistent with the data that it indexes. You can use the REBUILD INDEX utility to rebuild the indexes. Alternatively, you can use the RECOVER utility to recover catalog and directory indexes if the index was defined with the COPY YES attribute and if you have a full index image copy.
FL 500
If you recover the catalog and
directory to a point in time before the REORG operation that converted the DSSIZE value for SPT01
and SYSLGRNX to 256 GB, the DSSIZE value for those table spaces is reverted to 64 GB.