While journal and backup information can be used to recover page data files, you can encounter other failures from which you must be prepared to recover. Other files that can fail include: JOURNAL1, JOURNAL2, the redolog, or the contingency log. The following describes these situations.
The journals provide an audit trail of all changed physical pages. If one of your journals fails, you are unable to recover for subsequent page file failures. For this reason, you should immediately reset the continuous backup process with a new master backup.
After the Data Object Broker is started, use subsequent journal images, along with your backup, to recover any page file failures up to the time of failure. Although the page images contained within the failing journal are lost, your repository and other control files are still intact.
The journal has a two-fold function: as part of continuous backup, and as part of checkpoint processing. When acting as a part of continuous backup, the loss of the journal means that a new full system backup must be created. For checkpoint processing, the journal is performing a caching role; if it is damaged before the last checkpoint is fully propagated to the Pagestore, it can impact Pagestore integrity.