Share via


Design for a Clustered Instance of a Service or Application

Applies To: Windows Server 2008

By creating a failover cluster and configuring a service (such as Dynamic Host Configuration Protocol (DHCP)) or a server-client application in the cluster, you can help provide high availability for the service or application. This design does not include a disaster recovery option as part of the failover cluster, which makes this design less expensive and simpler to deploy. For additional designs, such as a design that includes a disaster recovery option, see Mapping Your Deployment Goals to a Failover Cluster Design.

Note that for the maximum availability of any server, it is important to follow best practices for server management—for example, carefully managing the physical environment of the servers, testing software changes before placing them into production, and carefully keeping track of software updates and configuration changes on all clustered servers.

A failover cluster usually includes a storage unit that is physically connected to all the servers in the cluster, although any given volume in the storage is only accessed by one server at a time. The following diagram shows a simple two-node failover cluster connected to a storage unit and supporting an application.

Failover on a two-node cluster supporting an application

To learn more about this design, see Example, Clustered Instance of a Service or Application.

For information about configuring this design, see Checklist: Clustered Instance of a Service or Application (https://go.microsoft.com/fwlink/?LinkId=129122).