An error-budget is a quantified tolerance for failure, defined as the complement of a service's reliability target (SLO) over a rolling time window. Calculated as (1 - SLO) multiplied by total operations, it is tracked through cumulative burn-rate metrics that signal when error consumption exceeds acceptable velocity. The budget enforces a trade-off: rapid error accumulation triggers release freezes until the budget recovers. [formal: error_budget | substrate: behavior | horizon: hours | explicit: yes | epoch: 0.01]
Accepted ontology entry
error-budget
An error-budget is a quantified tolerance for failure, defined as the complement of a service's reliability target (SLO) over a rolling time window. Calculated as (1 - SLO) multiplied by total operations, it is tracked through cumulative b…
Definition
Why it is in scope
A human-made accounting construct used in site reliability engineering: the quantified allowance of error or failure a service may accumulate within a defined period before violating its reliability target. Built to persist through SLO definitions, monitoring dashboards, and operational playbooks that translate abstract reliability goals into concrete tolerance thresholds.
Names and aliases
- error-budgeten · CANONICAL
Relations from this entry
- cms7wyk72004ph6s8d82lhwqlDEPENDS_ON →
Remove reliability — error-budget stops operating because it cannot compute tolerances without reliability targets (SLOs). Error-budget is computed from (1-SLO)*ops; remove the reliability concept and the calculation has no inputs. This is a present-tense dependency, not just historical association.
- cmrwiv1rn00a8soacg5vdpiogINSTANCE_OF →
error-budget is a specific kind of metric — a quantified measure used in SRE to track allowable error/failure accumulation against a reliability target. An error budget is calculated, tracked, and visualized as a metric value. Specific→general kind relation.
- cms7wyk72004ph6s8d82lhwqlSERVES →
An error budget is allocated and managed for the sake of maintaining reliability within acceptable failure thresholds. Law 8d: servant (error-budget) → master (reliability). Its designed purpose is to quantify how much unreliability is acceptable, thereby guiding decisions that preserve reliability — it exists to serve reliability.
- cmr9uz3vv00elhcxfruyltnd4DEPENDS_ON →
error-budget needs measurement to operate now: an error-budget is a quantified tolerance for unreliability, and it only functions when measurement tracks actual errors against the budget. Remove measurement and the error-budget stops operating — it becomes just a wish. Law 8 removal test passes.
- cmsinft2b00jqnobphwthgc0iDERIVED_FROM →
Error-budgets were derived from service-level-objectives: an error budget is the acceptable amount of failure derived by subtracting the SLO target from 100%. SLOs came first — error budgets are a direct consequence of SLO targets. Which existed first? SLOs. SLOs fed into error-budget.
Relations to this entry
No accepted relations in this direction.
Record identity
- Created
- Aug 7, 2026, 7:35 AM UTC
- Content hash
- 8522a6fbeb9217424e3d4e73fad1c796dd10411c537b090278f879e2d2665b8e