About the Uptime & SLA Calculator
This uptime and SLA calculator converts an availability target into the downtime it allows. Enter a percentage such as 99.9% (“three nines”) and it shows how many seconds, minutes or hours of outage that permits per day, week, month, quarter and year. Switch the mode to work backwards: enter the downtime you actually had in a period and it returns the availability you achieved and how many nines that represents.
It is built for SREs and engineering managers setting service level objectives and error budgets, founders writing an SLA into a customer contract, and buyers checking what a vendor’s promise really means. You can also enter how many services a request depends on in series — if each must be up for the request to succeed, overall availability is the product of the individual figures, which is always lower than any one of them.
Periods use an average Gregorian year of 365.25 days, so a month is 30.44 days (730.5 hours) and a quarter is 91.31 days. Real SLAs often measure calendar months, exclude scheduled maintenance and count only minutes when a service is fully unavailable, so check your contract’s definitions.
With the default inputs, the downtime per month is 43m 50s. Change any value above to recalculate instantly.
How to use the uptime & sla calculator
- 1Choose whether you want allowed downtime from an SLA or uptime from measured downtime.
- 2Enter the SLA percentage, or the downtime and the period it occurred in.
- 3In SLA mode, set how many services a request depends on in series (1 for a single service).
- 4Read the allowed downtime per day, week, month, quarter and year.
- 5Compare against the table of common SLA levels when choosing a target.
Formula and method
Availability is the fraction of time a service is up, so the downtime it allows is simply the remaining fraction multiplied by the length of the period. At 99.9%, 0.1% of a 730.5-hour month is 43.83 minutes; at 99.99% it is 4.38 minutes. Periods are based on a 365.25-day year: a day is 86,400 seconds, a week 604,800, a month 2,629,800 and a year 31,557,600.
When a request depends on several services that must all be up, their availabilities multiply, so three services at 99.9% give about 99.7% overall. In reverse mode, measured downtime is divided by the period length and subtracted from 1. The “number of nines” is −log₁₀ of the unavailability, so 99.9% is exactly 3 nines and 99.95% is about 3.3.
- A
- Availability of one service (e.g. 0.999)
- n
- Number of services in series
- Period
- Length of the measurement window in seconds
Worked examples
99.9% SLA (three nines)
0.1% of a 2,629,800-second month is 2,629.8 seconds, about 43 minutes 50 seconds. The same SLA allows 86.4 seconds a day, about 10 minutes a week and 8 hours 46 minutes a year.
99.99% SLA (four nines)
Four nines leaves 0.01% unavailability: about 4 minutes 23 seconds a month, 8.6 seconds a day and 52.6 minutes a year — little room for manual incident response.
45 minutes of downtime in a month
45 minutes is 2,700 seconds out of 2,629,800, or 0.1027% downtime, so availability was 99.8973% — just short of a 99.9% SLA.
Three 99.9% services in series
0.999³ = 0.997003, so the combined system is 99.7% available and the allowed downtime roughly triples to about 131 minutes a month.
Frequently asked questions
How much downtime is 99.9% uptime?+
99.9% availability allows about 43 minutes 50 seconds of downtime per month, 1 minute 26 seconds per day, and about 8 hours 46 minutes per year, based on a 365.25-day year.
How much downtime does 99.99% allow?+
Four nines allows roughly 4 minutes 23 seconds per month and about 52 minutes 36 seconds per year. At this level outages usually need automated failover because there is little time for a human to respond.
What is an error budget?+
An error budget is the amount of unreliability an SLO permits — 100% minus the target. A 99.9% monthly SLO gives a budget of about 43.8 minutes; teams spend it on deploys and experiments and slow down releases when it runs out.
What is the difference between an SLA, SLO and SLI?+
An SLI is the measured indicator (for example the share of successful requests), an SLO is the internal target for that indicator, and an SLA is the contractual promise to customers, usually with service credits if it is missed. SLAs are normally set looser than SLOs.
Why does availability drop when services depend on each other?+
If a request needs every service in a chain to be up, the chance that all are up is the product of their availabilities. Three services at 99.9% each give 99.7%, so the combined system has about three times the downtime of any single part.