Availability

Availability measures whether a service meets a defined success criterion during a measurement window. This lesson introduces the time-based calculation, translates targets into downtime, and explains why the measurement rules matter as much as the percentage.

What is availability?

Availability is the fraction of measured time or requests for which the service satisfies its availability criteria. Define what counts as usable: a reachable server may still fail to complete the user’s operation. A measured value of 100% means no failures under that definition in the observed window; it is not a guarantee about the future.

Measuring availability

For a time-based metric, availability A is calculated over an explicit interval using the following formula:

Availability targets are often described in “nines.” The table below converts time-based targets into downtime using a 365-day year, a 30-day month, and a 7-day week. Values are rounded.

The Nines of Availability