Microsoft Windows 2000 Server, and especially Advanced Server with its clustering support, provides an excellent environment in which to build a truly fault-tolerant system. Of course, avoiding the faults in the first place is even better than handling them once they've happened, but the realistic system administrator knows that a problem will occur sooner or later, and he or she plans for it. Chapter 34 covered disaster planning in depth, so you should refer to that chapter for information on how to prepare for major problems and how to build a full disaster recovery plan to quickly resolve them.
This chapter focuses primarily on the hardware and software tools that allow you to build a highly available and fault-tolerant Windows 2000 environment. Remember, however, that no matter what hardware and software you deploy, building and deploying for high availability and fault tolerance requires time, a clear understanding of the trade-offs required, and—most importantly—discipline. You'll need to make informed decisions about your real requirements and determine the resources available to meet those requirements. When planning for a highly available and fault-tolerant deployment, you should consider all points of failure and work to eliminate any single point of failure. Redundant power supplies, dual disk controllers, multiple network interface cards (multihoming), and fault-tolerant disk arrays (RAID) are all strategies that you can and should employ.