SQL Server Database Mirroring for SharePoint Farms

Introduced in SQL Server 2005 Service Pack 1, database mirroring is a software solution that delivers high availability/database redundancy of SQL databases, including SharePoint databases. SQL mirroring is highly valuable for SharePoint environments because it provides the promise of having an always-available constant replica of SharePoint databases on a different server and being able to fail over to that server, either automatically or manually, as needed.

There are two primary partners in a database mirroring session: the principal server and the mirror server. Essentially, database mirroring works by maintaining a copy of the same database or databases on both partners. This is accomplished by streaming the active transactional log records from the principal server to the mirror server, which in turn applies the log records to the mirror database. In comparison to failover clustering, database mirroring works on a per-database basis, and provides high availability and data protection for both storage and hardware failures. In addition to the two partners involved in a database mirroring session, a third server instance, known as the witness server, can be added to provide automatic failover by verifying whether the principal server is up and functioning.

Utilizing SQL database mirroring for a SharePoint environment can have the advantage of having a second, fully complete copy of all SharePoint content automatically replicated to a secondary location. This is the equivalent of running a constant backup of SharePoint data, and has the added advantage of allowing for instant failover to a secondary SQL server in the event of a failure. Indeed, this version of SharePoint has been written to be “mirror-aware,” and all databases can be configured with a secondary SQL server instance chosen to allow for mirroring failover scenarios.