Developer & Tech

Uptime SLA Calculator

What this does

Translate an availability percentage into the exact downtime budget per day, 30-day month and year, humanized as clock time.

Enter your details

Runs in your browser

Calculator inputs

Using the uptime sla calculator

  1. 01

    Enter the promised availability

    Take it straight from your SLA contract or internal SLO, e.g. 99.9.

  2. 02

    Read the budgets

    Per-day, per-month and per-year figures show how little (or much) failure time the number buys.

Nines escalate brutally

  • 99% (“two nines”); 3.65 days/year
  • 99.9%; 8.77 hours/year
  • 99.99%; 52.6 minutes/year
  • 99.999%; 5.26 minutes/year

Budgets drive design

Each additional nine typically demands redundant everything: multi-zone deployments, automated failover, canary rollouts. Knowing that 99.9% allows only 43 minutes monthly reframes “minor incidents”; one forgotten alert can spend the month’s entire budget.

The math behind this calculator

downtime = window × (1 − availability) windows: 86,400 s/day · 2,592,000 s/30-day month · 31,536,000 s/year

Availability is the fraction of time a service must answer. Multiply each calendar window by (1 − availability) to get the allowed outage seconds, then format them as mm:ss or hh:mm:ss so budgets feel concrete in on-call reviews.

Assumptions & limitations

  • 30-day month = 2,592,000 seconds; year = 365 days.
  • All partial outages count against the budget equally.
  • Scheduled maintenance counts or not per your contract; this shows raw math.

Worked example

A 99.9% SLO permits 86.4 seconds of downtime per day and exactly 43 minutes 12 seconds per 30-day month (2,592,000 × 0.001).

Frequently asked questions

Why use a 30-day month?
It is the industry convention (2,592,000 s) making months comparable; real months vary 28–31 days, so contracts specify which they mean.
Does planned maintenance count?
Contracts differ. This tool shows raw unavailability budgets; subtract agreed maintenance windows from the allowance yourself.
How do I measure actual availability?
Aggregate probe results over the window: successful requests (or seconds up) divided by total. Compare against this budget weekly.

Related calculators