A server is a 'central' computer that shares it's resources amongst many users. It benefits from being a more robust computer than the average desktop machine as it is generally expected to perform without significant interruption 24 hours a day, 365 days a year.

• Mirror the Server’s Disks. This is an absolute must. A disk failure in a server could take 2 to 3 days to recover from. A RAID (Redundant Array of Inexpensive Disks) utilises more than one disk so that in the event of a drive failure the system continues to operate as normal.
• Ensure your Backup software is backing up all of the server. Exchange and SQL Servers may require “agents” which are often overlooked.
A missing Agent could prevent recovery of your email system.
• Keep at least one set of backup disks (or tapes) off-site.
• When planning the size of your server disk storage allow for at least twice as much space as you think you will need and ensure the backup devices are large enough to accommodate this.
• Install an UPS and budget to replace it's battery every two to three years.
• Apply Service Packs regularly. Some service packs require a system reload, so if auto-update is used schedule a reload monthly or even weekly.
• Reboot older servers automatically on a regular basis to avoid 'memory leaks' which can slow or stop the server. This may also apply to newer servers running older applications.
• Monitor disk usage and never let the server run out of disk space. Lack of space will cause services to stop unexpectedly, in particular Email, and could cause corruption of critical system files.