How the numbers are worked out
Uptime is the share of time a service was working. The downtime a target allows is the rest:
allowed downtime = period length x (100 - uptime %) / 100For 99.9% over a 30-day month: 30 days is 43,200 minutes. 0.1% of that is 43.2 minutes, which is 43 minutes and 12 seconds.
The common targets
| Uptime | Per month (30 days) | Per year (365 days) |
|---|---|---|
| 99% | 7 h 12 min | 3 d 15 h 36 min |
| 99.5% | 3 h 36 min | 1 d 19 h 48 min |
| 99.9% | 43 min 12 s | 8 h 45 min 36 s |
| 99.95% | 21 min 36 s | 4 h 22 min 48 s |
| 99.99% | 4 min 19 s | 52 min 34 s |
| 99.999% | 26 s | 5 min 15 s |
Each extra nine cuts the allowance to a tenth. Going from 99.9% to 99.99% means the whole year's downtime must fit inside 53 minutes, including deploys that go wrong.
What the percentage does not tell you
- How it was measured. A check every 5 minutes can miss a 4-minute outage completely. A check every minute cannot.
- What counted as down. A slow page or a broken checkout may still count as "up".
- What was left out. Many service agreements exclude planned maintenance from the figure.
- How the downtime fell. One 43-minute outage and forty 1-minute ones give the same 99.9%, and feel very different to customers.