The conversion
Uptime percentage is the share of a period the service was available. Downtime is the rest. Multiply the downtime share by the period's minutes:
Downtime = (1 − uptime) × minutes in periodThe ladder of nines
- 99%7.2 hoursdowntime per month
- 99.9%43.2 minutesdowntime per month÷10 from 99%
- 99.95%21.6 minutesdowntime per month÷2 from 99.9%
- 99.99%4.32 minutesdowntime per month÷5 from 99.95%
- 99.999%26 secondsdowntime per month÷10 from 99.99%
The calculator's five presets on the same 30-day month, drawn as an ordinal ladder: the step multipliers are printed because the steps are not equal, and the rungs keep one uniform size because no spacing here claims a ratio. The table below carries the yearly figures.
| Uptime | Downtime/month (43,200 min) | Downtime/year |
|---|---|---|
| 99% | 7.2 hours | 3.65 days |
| 99.9% | 43.2 minutes | 8.76 hours |
| 99.95% | 21.6 minutes | 4.38 hours |
| 99.99% | 4.32 minutes | 52.6 minutes |
| 99.999% | 26 seconds | 5.26 minutes |
What the table tells you
- 99.9% still allows 43 minutes of monthly downtime — nearly a full work-hour outage every month is 'within SLA'.
- Each nine is ~10× harder. The jump from 99.9% to 99.99% eliminates 39 of those 43 minutes — and typically costs far more than 10× the engineering.
- Monthly vs yearly framing matters. '8.76 hours a year' sounds fine; '43 minutes a month' is the same number wearing honesty.
Uptime is not the whole story
A percentage says nothing about when downtime hits (3 AM vs Black Friday), how it's distributed (one outage vs fifty blips), or whether degraded-but-up counts. Define what 'up' means for your service — full functionality? Core path working? — before arguing about the nines.
Assumptions and limits
- Month = 30 days (43,200 minutes); year = 365 days. Leap effects ignored.
- Percentages are definitional; no vendor SLA claims are made.
Common questions
How much downtime does 99.9% uptime allow?
About 43.2 minutes per month, or 8.76 hours per year, on this page's stated basis — a 30-day month of 43,200 minutes and a 365-day year. That's why 99.9% — 'three nines' — still permits a noticeable monthly outage: nearly a full work-hour outage every month is 'within SLA'.
What is the difference between 99.9% and 99.99% uptime?
A factor of ten in allowed downtime: 43.2 vs 4.32 minutes per month. The jump eliminates 39 of those 43 minutes — and typically costs far more than 10× the engineering, because each nine is disproportionately harder than the last. Monthly framing matters too: the same downtime reads very differently as minutes per month than as hours per year.
Is 100% uptime possible?
No — 100% uptime is a marketing claim, not an engineering target. Every system has maintenance, deploys, and dependencies that can fail. A percentage also says nothing about when downtime hits, how it is distributed, or whether degraded-but-up counts. Plan for the nines you can afford and design recovery for the rest.
Published · Updated . Availability planning guides for small web teams. Models are planning tools, not guarantees.
Sources: Google SRE Book: Service Level Objectives — the SLI, SLO and SLA vocabulary and the availability-nines framing behind this page. Every figure above is arithmetic on the stated windows.
The next outage is a matter of when.
Be the team with the plan: a one-page runbook with named roles, severity levels, and the first-15-minutes sequence.