How much downtime is 99.9% uptime?

99.9% uptime allows 1 minute 26 seconds of downtime a day, 10 minutes 5 seconds a week, 43 minutes 50 seconds a month and 8 hours 45 minutes 57 seconds a year.

How much downtime is 99.99% uptime?

99.99% allows 8.6 seconds a day, 1 minute a week, 4 minutes 23 seconds a month and 52 minutes 36 seconds a year. Each extra nine cuts the allowance by a factor of ten.

How do you calculate uptime percentage?

Divide the time the service was up by the total time and multiply by 100. A service down for 45 minutes in a 30-day month was up 43,155 of 43,200 minutes: 99.896%.

What is a composite SLA?

The availability of a system that depends on several services. Multiply the availabilities of services that must all work: two at 99.9% give 99.8%. Redundant copies raise it: two copies at 99.9% give 99.9999%, if they fail independently.

How do you calculate MTBF?

MTBF, mean time between failures, is total uptime divided by the number of failures. A server up 8,750 hours with 4 failures has an MTBF of 2,187.5 hours. Availability is then MTBF ÷ (MTBF + MTTR).

Mean time to repair or recover: total downtime divided by the number of failures. Ten hours of downtime over 4 failures is an MTTR of 2.5 hours. Halving MTTR raises availability as much as doubling MTBF.

What is an error budget?

The downtime an SLO allows over its window: (1 − SLO) × window. A 99.9% SLO over 30 days gives 43.2 minutes. In Google’s SRE practice, releases pause when the budget is spent.

How fast you use the error budget relative to the SLO window. At 1 the budget lasts exactly the window; at 2 it is gone halfway through. Google’s SRE workbook builds its SLO alerts on burn rate.

Free calculator

Uptime calculator for SLAs, error budgets and MTBF

See how much downtime an SLA allows, turn an outage into an uptime percentage, work out the availability of a whole stack, and track an error budget.

Updated October 27, 2026

Your numbers

Uptime target (SLA)

Up to four decimals, for example 99.95.

Common targets

Runs in your browser. Nothing you type is sent or stored.

Allowed downtime per month

43m 50s

At 99.9% uptime, with an average month of 30.44 days.

Per day: 1m 26s
Per week: 10m 5s
Per quarter: 2h 11m 29s
Per year: 8h 45m 57s

Downtime by SLA

SLA	Per day	Per month	Per year
90%	2h 24m	3d 1h 3m	36d 12h 35m
95%	1h 12m	1d 12h 31m	18d 6h 17m
99%	14m 24s	7h 18m 17s	3d 15h 39m
99.5%	7m 12s	3h 39m 9s	1d 19h 50m
99.9%	1m 26s	43m 50s	8h 45m 57s
99.95%	43.2s	21m 55s	4h 22m 58s
99.99%	8.6s	4m 23s	52m 36s
99.999%	864ms	26.3s	5m 16s

99.9% allows 8h 45m 57s of downtime a year. One more nine, 99.95%, allows 4h 22m 58s.

Quick answer

How do you calculate uptime and allowed downtime?

Uptime (availability) is the share of time a service works: availability = uptime ÷ total time × 100. The downtime an SLA allows is (1 − availability) × the period, so 99.9% allows 43 minutes 50 seconds a month and 8 hours 45 minutes 57 seconds a year. Each extra nine divides the allowance by ten: 99.99% allows 4 minutes 23 seconds a month.

How to use the uptime calculator

Pick a mode
SLA for what a target allows, Downtime to turn an outage into a percentage, Stack for several services together, Budget to track an SLO, MTBF for availability from failures and repair time.
Enter your numbers
Availability takes up to four decimals, so 99.995% works. A stack takes up to six services, each with its own number of redundant copies.
Read the result
Downtime per period and the table of nines, the composite availability, or the budget left and how fast you are using it.
Share it
Copy the link to send the exact numbers to your team. Nothing is stored on our side.

How the calculator works

Availability is the share of time a service works. The downtime it allows is (1 − availability) × the period. The calculator uses a 24-hour day, a 7-day week and the average calendar year of 365.2425 days, leap years included; a month is a twelfth of that year (30.44 days) and a quarter a quarter of it. Many contracts measure a calendar month instead, so check how yours defines the period.

Services that must all work are in series: multiply their availabilities. Redundant copies, where any one keeps the layer up, are in parallel: 1 − (1 − a)ⁿ for n copies. Both formulas assume failures are independent and fail-over is instant, so treat the composite as a ceiling, not a promise.

An error budget is the downtime an SLO allows over its window. Burn rate compares the share of the budget used with the share of the window gone: at 1 the budget runs out exactly at the end of the window, at 2 it runs out halfway. MTBF and MTTR give availability as MTBF ÷ (MTBF + MTTR).

Formula

Allowed downtime = (1 − availability) × period
Availability = uptime ÷ (uptime + downtime)
In series: A = A₁ × A₂ × … × Aₙ
Redundant copies: A = 1 − (1 − a)ⁿ
Availability = MTBF ÷ (MTBF + MTTR)
Burn rate = (budget used ÷ budget) ÷ (time gone ÷ window)

Worked example

A stack: load balancer 99.99%, two app servers at 99.9% each, database 99.95%, payments API 99.9%.
Two redundant app servers: 1 − (1 − 0.999)² = 99.9999%.
The stack in series: 0.9999 × 0.999999 × 0.9995 × 0.999 = 99.84%.
Downtime a year: 0.16% × 365.2425 days = 14 hours 2 minutes.
The payments API alone accounts for 8 hours 46 minutes of it.

Downtime by SLA, from 90% to 99.999%

Availability	Per day	Per week	Per month	Per year
90%	2h 24m	16h 48m	3d 1h 3m	36d 12h 35m
95%	1h 12m	8h 24m	1d 12h 31m	18d 6h 17m
99%	14m 24s	1h 40m 48s	7h 18m 17s	3d 15h 39m
99.5%	7m 12s	50m 24s	3h 39m 9s	1d 19h 50m
99.9%	1m 26s	10m 5s	43m 50s	8h 45m 57s
99.95%	43.2s	5m 2s	21m 55s	4h 22m 58s
99.99%	8.6s	1m	4m 23s	52m 36s
99.999%	864ms	6s	26.3s	5m 16s

SLA, SLO and SLI: the difference

An SLI is the measurement, such as the share of requests that succeed. An SLO is the target for it, such as 99.9% over 30 days. An SLA is the contract with customers, with consequences, such as service credits, when you miss it. Set the SLO tighter than the SLA so you notice trouble before it costs money.

Why each nine costs more

Each extra nine divides the allowed downtime by ten. At 99.9% a month allows 43 minutes 50 seconds; at 99.99% it allows 4 minutes 23 seconds, so one incident that needs a person to respond can use the whole month. Higher targets therefore rest on redundancy and automatic fail-over rather than on faster humans.

Uptime and SLA questions

How much downtime is 99.9% uptime?: 99.9% uptime allows 1 minute 26 seconds of downtime a day, 10 minutes 5 seconds a week, 43 minutes 50 seconds a month and 8 hours 45 minutes 57 seconds a year.
How much downtime is 99.99% uptime?: 99.99% allows 8.6 seconds a day, 1 minute a week, 4 minutes 23 seconds a month and 52 minutes 36 seconds a year. Each extra nine cuts the allowance by a factor of ten.
How do you calculate uptime percentage?: Divide the time the service was up by the total time and multiply by 100. A service down for 45 minutes in a 30-day month was up 43,155 of 43,200 minutes: 99.896%.
What is a composite SLA?: The availability of a system that depends on several services. Multiply the availabilities of services that must all work: two at 99.9% give 99.8%. Redundant copies raise it: two copies at 99.9% give 99.9999%, if they fail independently.
How do you calculate MTBF?: MTBF, mean time between failures, is total uptime divided by the number of failures. A server up 8,750 hours with 4 failures has an MTBF of 2,187.5 hours. Availability is then MTBF ÷ (MTBF + MTTR).
What is MTTR?: Mean time to repair or recover: total downtime divided by the number of failures. Ten hours of downtime over 4 failures is an MTTR of 2.5 hours. Halving MTTR raises availability as much as doubling MTBF.
What is an error budget?: The downtime an SLO allows over its window: (1 − SLO) × window. A 99.9% SLO over 30 days gives 43.2 minutes. In Google’s SRE practice, releases pause when the budget is spent.
What is burn rate?: How fast you use the error budget relative to the SLO window. At 1 the budget lasts exactly the window; at 2 it is gone halfway through. Google’s SRE workbook builds its SLO alerts on burn rate.

Cite this tool

Free to link and quote. Use this reference:

SaaSCity, "Uptime & SLA Calculator", October 27, 2026. https://saascity.io/free-tools/uptime-calculator