Uptime and downtime calculator
Convert an uptime percentage into the downtime it allows, or find the uptime you delivered from the minutes you were down. It runs in your browser; nothing is sent anywhere.
Allowed downtime for an uptime target
A number from 0 to 100.
| Period | Allowed downtime |
|---|---|
| Per day | 1m 26s |
| Per week | 10m 5s |
| Per month | 43m 48s |
| Per quarter | 2h 11m 24s |
| Per year | 8h 45m 36s |
Uptime from downtime
Uptime: 99.8973%
Downtime at the usual targets
How it is calculated
Allowed downtime is the length of the period multiplied by the share of it you may be down: period × (100 − uptime) / 100. For 99.9% over a 30-day month that is 43 200 minutes × 0.001 = 43.2 minutes.
Calendar lengths differ, so here is the convention used above: a year is 365 days, a month is one twelfth of a year (30.4 days) and a quarter is a quarter of a year. Over a real 30-day month the figures are about 1.4% lower, and in a leap year the yearly figure is about 0.3% higher. Contracts usually state which they mean.
Uptime from downtime is the same formula the other way round: 100 × (1 − downtime / period).
Questions
How much downtime is 99.9% uptime?
About 43 minutes 48 seconds per month and 8 hours 45 minutes 36 seconds per year. This is the figure most SaaS service level agreements quote.
What is the difference between an SLI, an SLO and an SLA?
The SLI is what you measure, for example the share of requests answered successfully. The SLO is the target you set for it internally, for example 99.95% over 30 days. The SLA is the promise you make to customers, usually looser than the SLO and with a remedy such as service credits.
Is 99.99% uptime realistic?
It leaves about 4 minutes 23 seconds of downtime a month. A single manual response to an alert can use all of it, so reaching it takes redundancy and automatic failover, not just good intentions.
Does scheduled maintenance count as downtime?
It depends on the agreement. Many SLAs exclude announced maintenance windows; others count everything. Check the wording, and publish maintenance in advance so the exclusion is easy to defend.
How do you measure uptime in practice?
By checking the service at a regular interval from outside, ideally from more than one region, and dividing the successful checks by all checks over the period. A shorter interval catches shorter outages.