Calculate remaining error budget based on your SLO target and current failure rate.
An error budget is the amount of unreliability an SLO permits: budget = (1 − SLO) × time_period. For example, a 99.9% SLO over a 30-day period allows (1 − 0.999) × 30 × 24 × 60 = 43.2 minutes of downtime. Tracking consumption against this budget lets teams make data-driven decisions about when to slow down feature velocity in favor of reliability work.
errorBudget = (1 - sloTarget / 100) × timePeriodMinutes
(1 − 0.999) × 30 days × 1440 minutes/day = 43.2 minutes — this is the classic 'three nines' monthly budget cited in SRE literature.
Most SRE practices freeze non-essential feature releases and prioritize reliability work until the service is back within its SLO for the measurement window.
Time-based budgets suit availability SLOs; request-based budgets (bad requests / total requests) suit success-rate SLOs. Use whichever matches how your SLO is defined.
Burn rate is the speed at which you're consuming the error budget relative to the time elapsed in the period — a burn rate of 1x means you'll exhaust the budget exactly at period end; faster burn rates trigger earlier alerting.