Automated cluster caching facility failover
If the primary cluster caching facility (CF) fails, Db2 cluster services automatically attempts to restart it and fails over the primary role to the secondary cluster caching facility (assuming the recommended two-cluster caching facility setup).
- How many cluster caching facilities
are in the Db2®
pureScale instance
If there is a single cluster caching facility in the Db2 pureScale instance, the instance will be brought down. If the cluster caching facility failed because of a software failure, a group restart will be initiated automatically. If the cluster caching facility failed because of a hardware failure, the user needs to fix the problem, after which a group restart will be initiated automatically.
If there are two cluster caching facilities in the Db2 pureScale instance, which is the recommended setup, Db2 cluster services attempts to fail over the primary role to the secondary cluster caching facility. If the primary cluster caching facility failed because of a software failure, it will be restarted automatically and reintegrated as the secondary cluster caching facility. If the primary cluster caching facility failed because of a hardware failure, the user needs to fix the problem, and then it will be restarted automatically and reintegrated as the secondary cluster caching facility.
- What the state of the secondary cluster caching facility is
when the primary cluster caching facility fails
If the secondary cluster caching facility is in PEER state, Db2 cluster services will fail over the primary role to it.
If the secondary cluster caching facility is not in PEER state, the instance will be brought down. Db2 cluster services then initiates a group restart with the former secondary cluster caching facility now in the primary role.