Uptime SLA Calculator
Inputs
| Availability target | 99.9 % |
|---|
Uptime SLA Calculator
Convert a service-level agreement availability percentage into allowable downtime per year, month, week, and day.
Inputs
Availability Target
Results
Enter a value to see results.
Allowable Downtime
Downtime per Period
Uptime and SLA Availability
A service-level agreement (SLA) availability percentage states the minimum fraction of time a service is guaranteed to be operational. It is the standard metric used by cloud providers, network operators, and software-as-a-service vendors when committing to reliability. Converting this single percentage into concrete downtime budgets — per year, month, week, and day — reveals the practical engineering constraint each tier implies.
Formula
Where the availability target is expressed as a decimal — so 99.9% becomes 0.999 — the allowed downtime in any period of length is:
tdown=(1−A)×TThis calculator uses the following standard period lengths:
| Period | Duration () |
|---|---|
| Year | 31,557,600 s (365.25 days) |
| Month | 2,629,800 s (30.4375 days) |
| Week | 604,800 s (7 days) |
| Day | 86,400 s (24 hours) |
The year uses 365.25 days to account for leap years; the month is a twelfth of that.
Worked Example
A SaaS product commits to 99.9% availability (three nines):
Atyear=0.999=(1−0.999)×31,557,600=0.001×31,557,600=31,557.6 s≈8.77 hours per yearThat 8.77 hours is the total allowable outage budget for the entire year. If a single incident consumes 6 hours, only 2.77 hours remain before the provider is in breach.
The Nines Table
| Availability | Common name | Downtime/year | Downtime/month |
|---|---|---|---|
| 99% | Two nines | ~87.7 h | ~7.3 h |
| 99.9% | Three nines | ~8.77 h | ~43.8 min |
| 99.99% | Four nines | ~52.6 min | ~4.38 min |
| 99.999% | Five nines | ~5.26 min | ~26.3 s |
Each additional nine reduces the downtime budget by a factor of ten. The engineering investment required to reach that reduction typically more than doubles — five-nines availability demands redundant hardware, automatic failover, zero-downtime deployment pipelines, and rigorous on-call processes.
Limitations
The formula measures the calendar-time budget, not the probability of any particular outage. A service that takes one long 8-hour outage and a service that takes 100 five-minute outages have the same availability number over the year, but very different user experience and failure patterns. Modern SLA contracts therefore complement availability percentages with response-time commitments, error-rate budgets (the Google SRE error budget approach), and per-incident maximum outage duration clauses.
Some SLA agreements exclude scheduled maintenance windows from the downtime calculation, so a stated 99.9% may include planned maintenance not counted in the budget above. Always check contract terms when comparing SLAs across vendors.
Frequently Asked Questions (FAQ)
What does an SLA availability percentage mean?
A service-level agreement (SLA) availability percentage states the minimum fraction of time a service is expected to be operational over a given period. 99.9% availability means that of every 1,000 time units, at most 1 unit may be a failure. In practice this translates to about 8.77 hours of allowable downtime per year. SLA penalties typically apply when cumulative outage time exceeds the calculated budget.
What do "two nines," "three nines," and "five nines" mean?
The "nines" shorthand counts the number of 9s in the availability percentage. Two nines is 99% (about 87.6 hours/year downtime); three nines is 99.9% (about 8.77 hours/year); four nines is 99.99% (about 52.6 minutes/year); five nines is 99.999% (about 5.26 minutes/year). Each additional nine reduces allowable downtime by a factor of ten, making each step progressively more expensive to achieve and maintain.
Does the SLA budget include planned maintenance windows?
This depends on the specific SLA contract terms. Some agreements exclude scheduled maintenance from the downtime calculation, effectively providing a higher engineering budget. Others count all non-operational time regardless of cause.
When comparing SLA tiers across providers, check whether planned outages are excluded, how outages are measured (often through synthetic monitoring), and what the measurement and reporting period is (monthly vs. yearly).
How is service availability typically measured?
Availability is usually measured through a combination of uptime monitoring, synthetic health checks, and error-rate tracking. A service may be "up" in the sense that its process is running but still failing to serve requests.
Modern SLAs often use error budgets based on the Service Level Objective (SLO) error rate rather than simple binary up/down monitoring. The Google Site Reliability Engineering book popularized this approach, where an error budget is 1 − SLO expressed as a fraction of total requests in a window.