Uptime SLA Calculator
Convert an uptime percentage into allowed downtime.
For example 99.9 for three nines.
Results update as you type.
How it works
An uptime SLA calculator converts an availability percentage, the kind quoted in a service level agreement, into the concrete downtime it permits over a year, a month, and a day. Percentages like 99.9 percent sound reassuring but are hard to picture, so this tool translates them into minutes of allowed outage, which is far easier to plan against and to hold a provider accountable to. This matters on both sides of an agreement. If you run a service, the figure tells you how tight your monitoring, redundancy, and on-call response must be to meet the promise. If you depend on a service, it reveals exactly how much disruption the SLA actually tolerates before penalties apply. The jump between tiers is dramatic: three nines allows over eight hours of downtime a year, while five nines permits barely five minutes. The formula works because downtime is the share of time the service is not available, which is one minus the uptime fraction, multiplied by the length of the period. Real agreements may exclude scheduled maintenance or measure over specific windows, so read the fine print alongside this estimate. Site reliability engineers, IT managers, and anyone negotiating or auditing an SLA use an uptime calculator to make availability targets tangible.
downtime = (1 − uptime%) × period
Uptime and downtime always sum to one hundred percent, so the allowed downtime fraction is one minus the uptime fraction. Multiplying that fraction by the length of a period gives the outage budget for it: a full year is 365 × 24 × 60 minutes, a month is that year divided by twelve, and a day is 24 × 60 minutes. The method works because downtime is simply the leftover share of time the service may be unavailable.
Source:Standard networking
We review all our tools from time to time to ensure accuracy.
Worked example
How to use this calculator
- 1
Enter your values
Enter target uptime.
- 2
Adjust the options
Fine-tune any options to match your situation.
- 3
Read your result
Downtime per year updates instantly as you type — no button to press.
Frequently asked questions
What does 99.9% uptime actually allow?
Three nines, or 99.9 percent uptime, permits about 525.6 minutes of downtime per year, which is roughly 8.76 hours, or about 43.8 minutes each month. It sounds nearly perfect, but that is still enough time for several meaningful outages annually. Seeing the figure in minutes helps you judge whether a given SLA tier truly meets your tolerance for disruption or only appears to on paper.
How much downtime does each extra nine remove?
Each additional nine cuts allowed downtime by a factor of ten. Two nines (99%) permits about 3.65 days a year, three nines about 8.76 hours, four nines about 52.6 minutes, and five nines only about 5.26 minutes. The cost and engineering effort to reach each higher tier rises steeply, so match the target to what your users genuinely need rather than chasing nines for their own sake.
Does scheduled maintenance count against the SLA?
It depends on the agreement. Many SLAs exclude planned maintenance windows announced in advance, counting only unplanned outages against the uptime figure. Others count all unavailability. This distinction matters, because a provider with generous maintenance exclusions may deliver less real availability than the headline number suggests. Always read how the SLA defines downtime and what it excludes before comparing two providers' promises directly.
Why show downtime per year, month, and day?
Different planning horizons need different views. The yearly figure is how SLAs are usually headlined, the monthly figure often matches the billing and reporting cycle providers use for credits, and the daily figure helps you gauge the impact of a single bad day. Showing all three lets you connect an abstract percentage to the operational reality of monitoring and incident response at each timescale.
How is downtime calculated from the percentage?
The calculator subtracts your uptime percentage from 100 to get the downtime percentage, converts it to a fraction, and multiplies by the minutes in each period. A year is treated as 365 days, a month as one twelfth of that year, and a day as 24 hours. This gives the maximum minutes the service may be unavailable while still meeting the stated availability target.