Windows Clustering

Consider clustering in your enterprise network storage strategy when you need greater availability with simplified management. Clustering reduces downtime by providing an architecture that keeps systems running in the event of a single server failure.

With Windows Clustering, you can connect two or more servers to form a cluster of servers that work together as a single system. Each server is called a node; each node can operate independently of the other nodes in the cluster. The Windows 2000 built-in clustering capability is based on open specifications, industry-standard hardware, and ease-of-use requirements.

Each node has its own memory, system disk, operating system, and subset of the cluster's resources. With a process called failover, if one node fails, the other node takes ownership of the failed node's resources. The cluster server then registers the network address for the resource on the new node so that client traffic is routed to the available system, which now owns the resource. When the failed resource is later brought back online, you can configure the cluster server to redistribute resources and client requests appropriately. A standard Windows 2000 cluster setup is shown in Figure 19.4.

Enlarge figure

Figure 19.4 Typical Two-Node Cluster Setup

The Clustering Service provides the following advantages:

Common Administration   With Cluster Administrator MMC snap-in, you administer a cluster as a single system. Also, a client computer interacts with a cluster as though it is a single server.

Load Balancing   Within the cluster, you can manually balance processing loads or unload servers for planned maintenance without taking data and applications offline.

High Availability   Clustering provides high availability by automatically recovering critical data and applications from many common types of failure. If a node in the cluster fails, Windows Clustering detects the failure and performs recovery of the processes active at the time of the failure. The failure of one node in the cluster does not affect the other node.

If clustering is not already part of your network and if you want to deploy Windows Clustering, you need to address several issues in your Windows 2000 Server deployment planning phase with regard to setting up a cluster environment. For more information about planning your cluster environment, see "Windows Clustering" in the Microsoft® Windows® 2000 Server Resource Kit Distributed Systems Guide, and also see Windows 2000 Advanced Server Help.

Note

For clustering solutions, use only certified configurations that are documented on the hardware compatibility list (HCL), which you can access online. For more information about this list, see the Microsoft Windows Hardware Compatibility List link on the Web Resources page at http://windows.microsoft.com/windows2000/reskit/webresources.

Considerations for Using Clustering in Your Storage Strategy

Consider the following advantages of incorporating a cluster environment into your storage planning:

• Clustering provides high availability without data replication, thereby providing data consistency without greatly impacting storage requirements and network traffic volumes.
• Clustering provides the ability to easily recover from software failures.
• Servers share a multiported disk array. Hardware RAID controllers provide the best performance to the external disk array.
• Clustering provides high availability of data but does not protect data integrity.

© 1985-2000 Microsoft Corporation. All rights reserved.