If the application requires uninterrupted access to data, the High Availability (HotStandby) feature in solidDB® can be used to provide high availability. All HotStandby features, such as failovers in the solidDB® server, multiple durability semantics, and read-on-standby are applicable to Universal Cache high availability setups. When you use HotStandby, the operation of Universal Cache is protected against single failures in the cache tier.
Important: When designing deployments with HotStandby, remember to plan how failure scenarios are handled in your environment. See “Failure handling in Universal Cache” for information about how failures are handled in the Universal Cache system.
Universal Cache deployment with backend high availability
Applications can have additional high availability or disaster recovery requirements on the backend data server as well. Different data servers implement such features using different technologies. The backend data server is still a single logical entity, but data and running processes can be distributed across multiple nodes.
In a typical backend high availability case, the IBM Infosphere CDC instances will run on the primary backend data server node. However, in case of the backend failure, the IBM Infosphere CDC instances will have to be restarted on a different machine as a part of the complete backend data server failover. The subscriptions will also have to be reconfigured to reconnect the relevant instances.
