Most arguments about monitoring turn out to be arguments about words, because uptime, response time, latency and availability all get used as if they meant the same thing, and they do not. The guides in this section define the terms a monitoring setup is built on, and say what each one is measuring.
The vocabulary splits into three groups. First, the numbers: what uptime means, how the percentage is calculated, what a string of nines allows in real minutes of downtime, and how time to first byte, response time and network latency divide one request between them. Second, the mechanics: what server monitoring covers beyond a reachability check, why a monitoring system rechecks a failed result from other locations before it believes it, and how synthetic checks differ from timings collected out of real visitors' browsers. Third, the terms that describe outcomes rather than measurements: what a DNS blacklist listing does to your email, what belongs on a status page, and why an automated check is worth more than an occasional look at the site.
Each guide answers its question in the first paragraph, then gives the working detail underneath, with the concrete formula or command where one exists. They are written to be read on their own, so where two terms are easy to confuse the guides link across to each other rather than repeating the explanation.
The definitions come out of running the checks. HostTracker has monitored websites since 2004 and now watches over 500,000 websites from 300+ checkpoints in 158 cities, across 13 monitor types, sending alerts by email, SMS, voice call, Slack, Telegram and more when a result changes. To see how the terms below map onto actual checks, start from the feature set.