Implementing Failover Clusters

Summary

As you’ve seen, failover clustering gives you the ability to run a virtual server on top of another system. If the physical system hosting the virtual server fails, another system in the cluster will assume control of the virtual server. For clients, this provides a high level of reliability to crucial data. Now you have a server that isn’t dependent on any one physical system to run.

With solutions available on both Windows and Linux and plenty of industry pressure to offer more failover clustering solutions, it’s likely that clusters will become easier to deploy and much more common in the coming years. Another form of clustering, load-balanced clustering, provides a different level of virtualization than with failover clustering. With load-balanced clustering, a virtual server can be hosted by up to 32 nodes simultaneously, providing for a way to handle a high volume of traffic that would normally be impossible for a single server to handle. You’ll look at this form of virtualization next in Chapter 10.