How to calculate uptime from incident logs

If you keep an incident log, you already have everything needed to calculate uptime.

The formula

uptime % = (total time − downtime) ÷ total time × 100

Step by step

1. Pick the period. For example, a 30-day month has 43,200 minutes.
2. List each incident’s duration. Use the time from when users were affected until they were not, not the time until the postmortem was written.
3. Add them up. Suppose you had 12, 7 and 21 minutes — a total of 40 minutes.
4. Apply the formula. (43,200 − 40) ÷ 43,200 × 100 = 99.907%.

Decisions that change the answer

• Do planned maintenance windows count?
• Do partial outages count in full, or weighted by affected users?
• Do you measure from your own monitoring or from the customer’s side?

Write these rules down before you start calculating, and apply them every month in the same way. Consistency matters more than which rule you choose.

Reverse it

You can also work backwards: use the calculator to see how many minutes your target allows, then compare that with your incident log to see how much error budget is left.