Microsoft SQL 2005 Clustering Load-Balancing Definition

Microsoft SQL Server 2005 has support for clustering. Simply put, clustering is the use of two or more computers used together to help prevent downtime on your SQL servers.

Failover Clustering

A failover cluster works by having multiple nodes, or servers, with identical installations of SQL Server. When the primary server becomes unavailable for some reason, the resources from the primary server are switched to another node in the cluster group, providing continuous access to your users.

Load-Balancing Clustering

Another option for clustering is to use Network Load Balancing. This type of cluster operates differently. Multiple nodes are still set up, but they run independently. Requests from your users are divided between the servers to reduce the work load on each one. If a server fails, the requests are simply distributed to a different server.

Clustering Background

Failover clustering has been a built-in availability feature in SQL Server since SQL Server version 6.5. SQL Server 2005 offers a higher degree of support and more options for your clustering. Clustering is available in the Enterprise and Standard versions of SQL, with a limit of two-node clustering for the database server in the Standard version.

References

• SQL Server Performance: Does SQL Server 2005 Clustering Really Make Sense for My Organization's Needs?
• Microsoft: SQL Server 2005 Failover Clustering White Paper
• Microsoft: SQL Server 2005 Features Comparison