Use either of the following methods to minimize or prevent data loss on your servers:
- Install an uninterruptible power supply.
- Take steps to minimize the effect of any single point of failure.
Use an uninterruptible power supply
One of the leading causes of system failure is the interruption of power supply. When you use an uninterruptible power supply, this enhances fault tolerance in your Exchange 2000 organization. An uninterruptible power supply is a necessity for servers that contain mission-critical data, especially in large enterprise server environments. An uninterruptible power supply is a device that allows your computer to keep running for at least a short time when the primary power source is lost. Such a device also provides protection from power surges. If your servers need heating or cooling to keep the hardware working properly, also consider providing a fault-tolerant climate control system (for example, battery backup power for each heating or cooling unit).
Minimize single points of failure
You can increase the fault tolerance of your Exchange Server organization by incorporating additional hardware that duplicates the hardware resources used in your server environment. Such hardware configurations permit one path of input/output data, or one physical hardware component of a server, to fail without affecting the operations of the server. Examples of such components include Microsoft Windows 2000 Server clustering components, redundant array of independent disks (RAID) components, network components, and storage area network (SAN) components.
The additional hardware configurations that you should implement to minimize single points of failure in your server environment vary, depending on what type of components that you want to make resistant to hardware failure. Such hardware is typically included as part of a turnkey solution provided by hardware manufacturers. The use of such hardware is an important fault-tolerance measure, especially in large enterprise server environments. Some hardware manufacturers even provide products that back up and restore data specifically on servers running Exchange 2000.
For more information about how to implement this hardware in your organization, contact your hardware manufacturer.
Note To minimize downtime costs, including sales and productivity losses, keep replacement hardware available for your production servers. For example, you can save your company time and resources if you have replacement network cards on hand to help you troubleshoot the cause of a problem. Replacement hardware should include alternate backup servers, network adapters, video and hard disk controller cards, routers, cables, hard disks, and power supply components.
For more information, visit the following Microsoft Web site: