RPID overview
The RPID function can be used to analyze the RECON data sets and locate valid recovery time spans for one or more databases without the need to manually compare DBRC histories for each database.
Recovery time span (RTS)
A recovery time span (RTS) is a period of time during which a database is not allocated and, therefore, the database can be recovered. Although valid recovery times might exist during which the database is allocated, RPID does not currently analyze log records and therefore, does not report on these recovery times.
The following sequence of events illustrates the concept of a simple RTS:
- Event 1:
- Database X was started with the /STA command at Time 1:
- Event 2:
- Database X was unallocated with the /DBR command at Time 2:
- Event 3:
- Database X was started with the /STA command at Time 3:
Time Span => 21.320 08:54:16.730541-0800 to 21.320 08:59:01.75235-080021.320 08:59:01. 75235-0800
The sequence of these three events shows that:
- Database X was successfully started and allocated at Time 1
- The database remained allocated until it was unallocated at Time 2
- Database X was successfully started again at Time 3
This sequence shows that an RTS exists between Time 2 and Time 3. The database can be recovered successfully to any timestamp within this RTS.
Timestamp recovery (TSR)
RPID analyzes database allocation records in the RECON data sets to identify RTSs to which one or more databases can be recovered by using timestamp recovery (TSR). RPID also analyzes database recovery records in the RECONs to identify recovery point timestamps to which individual databases can be recovered by using TSR.
Recovery point
A recovery point is a timestamp that can be used as input to IMS Database Recovery Facility, or in some cases to other database recovery utilities such as, the standard IMS Recovery utility (DFSURDB0), as a time to which a database can be restored.
- A batch image copy (IC) run
- An online IC run (in cases where the databases are not truly allocated)
- A complete change accumulation (CA) run
- A recovery run (Full or Partial), or an offline reorganization run
HALDB IMS OLRs and IMS Online Reorganization Facility
The run times of HALDB IMS Online Reorganizations (IMS OLRs) and IMS Online Reorganization Facility are also analyzed.
For IMS OLR and IMS Online Reorganization Facility, RPID indicates whether the databases were truly allocated and therefore, whether the run time is a valid TSR point.
HALDB partitions that have been part of an IMS Online Reorganization have two sets of DBDSs, that are known as A-J and M-V.
At any point in time, either set might be active and the other set might be inactive (an IMS OLR is not in progress), or either set might be a shadow (an IMS OLR is in progress).
Both sets of DBDSs are recognized as one logical partition when RPID analyzes their associated RECON records.
Recovery points available for recovery
- Times of complete change accumulation runs.
- Times of batch image copy runs, and times of online image copy runs when the databases are not allocated.
- Times of offline reorganizations and online reorganizations when the databases are not allocated. In the second case (OLR), RPID indicates whether there was a subsequent image copy during the recovery time range.
- Times of database recovery, either full or partial recoveries.
Conditions and dependencies
- To manually create a recovery time span or recovery point for a TSR for all of the databases in a group, run the Recovery Point Create (RPCR) function that is described in Recovery Point Creation function (RPCR).
- The RPID function requires that you specify the target databases by using the DBLIST supporting function.
- The RPID function can use the RECON copy data sets that are created by the RECONCOPY
supporting function, or the data sets identified on the ALTRECONS supporting function.Note: Use either RECONCOPY or ALTRECONS in conjunction with RPID to prevent input/output (I/O) to the live RECON data sets.