Full System Recovery: DB2 MultiNode Agent
A full system recovery of a DB2 client involves the following phases:
Before you begin a disaster recovery, make sure to perform the following:
- Perform frequent DB2 MultiNode backups and the latest file system.
- Make sure that all the DB2 database applications, DB2 DPF or DB2 pureScale instances, and the existing databases for the DB2 DPF or DB2 pureScale instances are running when you perform a full backup. If these applications are not running, you must restore them manually after recovering the file system.
- Make sure that the install directories and appropriate application files have been backed up by the File System Agent.
- On Windows, make sure that the system state has been backed up.
As a best practice, it is recommended that you group databases into multiple subclients as follows:
- Add larger databases into separate subclients.
- Add small databases together into one or more subclients. This is important for the following reasons:
- During Disaster recovery, when you have to quickly rebuild the entire instance, you can concurrently restore all the subclients together.
- During Backup failures, the backup will restart from the beginning of the database instead of from the beginning of the entire instance. Similarly, this will ensure that large database backups are not affected by restarts from a smaller database.
- Once the subclients are created, schedule frequent backups for dynamic data and regular backup schedules for static data in the DB2 Server.
Recovering the DB2 Database
To recover the DB2 MultiNode database in the event of a disaster, perform the following tasks:
- Rebuilding the Operating System.
- Restoring the DB2 MultiNode Database to the same or different client computer.