metric

Error Budget

The amount of unreliability an SLO permits, treated as a resource that feature velocity spends.

sreslorelease-policy

If the SLO is 99.9%, the error budget is the remaining 0.1%. Spending it on a risky deployment is a legitimate choice; spending it accidentally on an outage is the same withdrawal from the same account.

Its real function is organisational rather than technical. It converts the recurring argument between "ship faster" and "make it more reliable" into a policy: while budget remains, ship; when it is exhausted, feature work pauses and reliability work takes priority until it recovers. Both sides agreed to the number in advance, so the decision is not relitigated during an incident.

An error budget that is never consumed is a sign the SLO is set too conservatively and the organisation is paying for reliability nobody asked for.