The error budget burn rate calculator gives the error budget of a service level objective, the burn rate of the current error rate, the share of budget used per hour and the hours before the budget is exhausted.
How the calculation works
- Enter the SLO and the compliance window.
- Enter the current error rate and the budget already consumed.
- The calculator gives the burn rate, the error budget, the hourly consumption, the hours left and the alert thresholds.
Example
A 99.9 % SLO over 30 days allows 0.1 % of errors, which is 43.2 minutes of complete outage. With an error rate of 1.44 %, the burn rate is 1.44 / 0.1 = 14.40: the budget burns 2.00 % per hour. If 20 % is already used, the remaining 80 % last 40.0 hours. The alert thresholds are 14.4 for 2 % in 1 hour and 6.0 for 5 % in 6 hours.
Why it matters
The calculator helps site reliability engineers, developers and managers who define service level objectives to set alerts that react quickly to real incidents without waking people for minor blips.
Good to know
Alerts on the burn rate are better than alerts on a fixed error threshold, because they take into account how much budget the SLO leaves. Teams often combine a long window and a short one, for example one hour and five minutes, so that an alert fires only when the problem is both significant and still happening. The error rate here can mean failed requests or degraded minutes, depending on how the SLO is defined, so use the same unit as the objective. Higher objectives leave very small budgets: 99.99 % over 30 days is only 4.3 minutes. A budget that is never spent can also be a sign of excess caution.
Related tools
The uptime and SLA calculator converts availability targets into downtime, the MTBF and availability calculator links failures and repairs, the queue wait time calculator covers waiting lines and the server capacity calculator sizes servers.