Monitoring

Monitoring is the continuous, automated observation of network equipment, servers, switches, and other IT infrastructure, usually through regular checks of metrics such as reachability, load, or temperature. It exists because many problems, such as a hard drive filling up or a switch overheating, show warning signs days or weeks before an actual failure, but only get noticed if someone or something checks regularly.

In practice, a monitoring system runs in the background, polling servers, switches, the UPS, and other devices at fixed intervals, and automatically sends an alert by email or SMS whenever a threshold is crossed, such as high temperature in the server rack or an internet line going down. Results are often viewable as historical trends, making it easier to spot gradual changes.

A common mistake is monitoring that is technically set up but whose alerts nobody checks regularly, or that produces so many low-priority alerts that the important ones get buried. Monitoring alone also does not replace a response: an alert nobody acts on prevents no failure.