Roll-Forward Recovery of a Standby Database

It may be desirable to maintain a copy of your database on another machine in standby mode. The standby copy of the database can be made quickly and easily, by backing up the database and then restoring and recovering it on another machine.

As transaction journals become archived (or offline) on your primary system, you can then copy them to the standby system and apply them to the database.

When a transaction journal switch occurs, a journal switch control record is written to the end of the journal. When this record is processed by roll-forward recovery and it is the final record in the transaction journal, the recovery process knows that later transaction journals can exist and the state is preserved so that a further roll-forward recovery from that point is possible.

When the batch JADE Database utility (jdbutilb) is run to recover the database, it processes the next journal and whatever later journals exist in the directory. However, on encountering a journal switch control record, the online JADE Database utility (jdbutil) waits for the next journal file to be provided or for the user to finish or abort the recovery processing. For more details, see "Using the Restore Database Command" and "Using the Initiate Recovery Command", in Chapter 1.

The online and batch versions of the JADE Database utility support continuing roll-forward recovery over multiple process executions. Any other open action of the database cancels the preserved state that enables continuous roll-forward recovery.