All guides
Guide

Website uptime monitoring, explained simply

6 min read Updated 2026-08-11

What uptime actually means, how it's measured, and why "99.9%" hides more downtime than you'd think.

"Uptime" gets thrown around in marketing copy like it's a single number that tells you everything. It isn't. Understanding what uptime really measures — and what it quietly leaves out — helps you read status pages and SLAs without being misled.

What uptime actually is

Uptime is the percentage of time a service was reachable and responding correctly over a given window. It's measured by repeatedly making requests to the service and recording whether each one succeeded. The ratio of successful checks to total checks, expressed as a percentage, is your uptime figure.

Why 99.9% isn't as good as it sounds

Three nines — 99.9% — sounds impressive until you do the math. Over a year, 99.9% uptime allows for about 8 hours and 45 minutes of downtime. 99.99% ("four nines") drops that to roughly 52 minutes. 99.999% ("five nines") is about 5 minutes a year — and is genuinely hard to achieve. When a provider quotes an uptime number, always ask: over what window, and what counts as "down"?

Infographic comparing uptime tiers: 99.9% allows about 8 hours 45 minutes of downtime per year, 99.99% about 52 minutes, and 99.999% about 5 minutes
The "nines" ladder: each extra nine cuts the allowed annual downtime by roughly 90% — but costs exponentially more to achieve.

What gets excluded

Most SLAs exclude scheduled maintenance windows, force-majeure events, and problems caused by the customer's own code or configuration. That means the headline number can look great while real users still experience outages. The exclusions are usually fair, but they're worth knowing when you compare two providers.

Check frequency matters

If you only check once an hour, a 5-minute outage might fall entirely between two checks and never get recorded — your uptime log will show 100% even though real users were affected. Shorter intervals catch more, but cost more. For most sites, checking every 1–5 minutes is a reasonable balance.

Timeline diagram showing monitoring check dots along a time axis, with a short red outage gap falling between two checks and going undetected
A short outage can fall entirely between two checks — the sparser the checks, the more downtime goes unrecorded.

How to use this site for lightweight monitoring

CheckItsDown is built for on-demand checks rather than continuous monitoring, but the per-site history page gives you a quick read on recent reliability. Bookmark a site's page, hit "Check again" periodically, and watch the history list — a string of ups with the occasional down tells a very different story than a wall of red.

Infographic of a monitoring history log with a vertical stack of check entries, green dots for Up and occasional red dots for Down
A per-site history log: mostly green with the occasional red entry is normal; a wall of red is a sustained problem.

Keep reading