Error budget
Also known as Reliability budget
Definition
An error budget is the amount of unreliability permitted by a service level objective over its measurement window. If the objective is 99.9% availability, the budget is the remaining 0.1% of allowed unavailability under the defined measurement rules.
Turn an SLO into a tradeoff
An SLO defines the service level a team wants. The error budget is the distance between that target and perfect service. It gives product and engineering a shared way to discuss how much unreliability the service can absorb while still meeting its objective.
Suppose an illustrative service has a 99.9% monthly availability objective. Under a simple availability calculation, the allowed unavailability is 0.1% of the measurement window. The exact minutes depend on the window and exclusions.
Use the budget as feedback
If a team spends its budget quickly, it may pause risky changes, improve testing, strengthen rollback, or invest in capacity. If the service remains well within its target, the team may have room to move faster or accept a carefully evaluated experiment.
The choice is a policy decision. The budget does not make a failure harmless, and a budget that is never used does not prove the objective is well chosen.
Keep the calculation visible
Document the SLI, target, period, exclusions, and source. A change in monitoring can make the budget appear healthier without changing user experience.
Connect budget consumption to incident impact and the changes that preceded it. This makes the discussion more specific than asking whether a team has been reliable in general.
How this relates to Weave
Weave helps connect delivery and quality activity to reliability discussions. When an error budget is being consumed, teams can inspect recent changes, review patterns, and rework alongside service telemetry. The budget calculation itself belongs to the service measurement and SLO system.
Explore Engineering intelligence