This topic has not yet been written. The content below is from the topic description.
Database failures is one of the least often addressed areas when planning disaster recovery, and with good reasons. Relational databases are very tolerant of most environmental hiccups – but problems can occur. Because of this, most relational databases use resilient storage techniques for their storage (highly recommended). This allows us to migrate have separate database instances using the same storage which gives us two versions of the same database. This provides database resiliency. To handle this from an application perspective, we need to set up the databases to have a Virtual IP address – and the active database will have this address. With this approach, our application servers can connect simply to the virtual IP. In the event of a database failure, the second database simply assumes the virtual IP. No change is needed to the application server since the IP database IP (from the application server perspective) is not changing. Additionally, the backup database is up to date since it referencing the same storage as the previous hot database was using.