Problem solve Get help with specific problems with your technologies, process and projects.

How does failover clustering work in SQL 2005?

Learn how the failover clustering functionality has been extended in SQL Server 2005.

How does failover clustering work in SQL 2005? And how does it compare with the failure support in Oracle and DB2?

The failover clustering functionality in SQL Server 2005 has been extended in various ways.

First of all, SQL Server 2000 running on Windows 2000 Datacenter was limited to a four-server cluster. SQL Server 2005 running on Windows 2003 Server can now support up to eight nodes (depending on the edition of Windows 2003.)

Failover is now also supported for a wider variety of services. Analysis Services, Notification Services, replication, and SQL Server Agent are all cluster-aware in SQL Server 2005. This is a major factor for those organizations that utilize these features and need to maintain high availability!

But the biggest improvement in terms of database availability is not from the failover clustering side at all. A new feature called "Database Mirroring" will prove to be a much more interesting way of maintaining uptime.

This feature can be thought of as a more real-time form of log shipping. Transactions are broadcast continuously from the active node to a waiting database on another server that is constantly in a state of recovery. When or if the active node goes down, the mirror can take over almost instantly. This is a huge improvement over the lag -- sometimes several minutes -- that can be experienced with failover clusters.

More information on this and other features of interest to DBAs, can be found in the TechNet article, Overview of SQL Server 2005 Beta 2 for the DBA.

Dig Deeper on Microsoft SQL Server 2005

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.