Skip to content

reliability

Uptime

Uptime — the percentage of time a service is responding correctly to its consumers, typically measured by a continuous synthetic monitor.

Also known as: availability

Definition

Uptime is the share of time a service is healthy by some agreed definition (responds quickly, returns correct data, asserts pass). It is the headline number on most status pages and the input to nearly every operational SLO.

In Maxoperf

Maxoperf synthetic monitoring and status workflows use uptime as one of the signals that explain whether a service stayed reachable before, during, and after a release window. Pair it with latency and assertion checks so “up” means useful, not merely reachable.

Common pitfalls

  • Quoting uptime without saying what counts as up — 200 OK is not the same bar as 200 OK with the right JSON shape in under 500 ms.
  • Reporting on single-region uptime — a regional outage in your probe provider can look like an outage in your service.

FAQ

What does 99.9 % uptime mean?

It allows roughly 43 minutes of downtime per 30-day window — beyond that, your error budget is exhausted.

How does Maxoperf measure uptime?

Each region of every monitor produces an outcome per check; uptime is the share of `up` outcomes over the configured rolling or calendar window.