What is an error budget?

WHAT THIS TESTS This evaluates whether you understand error budgets as the mechanism that aligns developers and operators around a single objective, rather than just a derived percentage.

A GOOD ANSWER COVERS An error budget is the maximum amount of unreliability a service is allowed to accumulate over a window while still meeting its SLO. It is calculated as one hundred percent minus the SLO target. For a 99.9 percent availability SLO, the budget is 0.1 percent of requests, or equivalently a small allowance of downtime per window. Its primary purpose is to reframe the eternal tension between shipping features and maintaining reliability as a shared, quantifiable resource. Both the product and SRE teams agree in advance: as long as the budget has room, the service is reliable enough and the team can spend velocity on new features and risk. Once the budget is exhausted, an error budget policy kicks in, typically freezing or slowing risky launches and redirecting effort to reliability, hardening, and reducing toil until the budget recovers.

COMMON WRONG ANSWERS Describing it merely as the inverse of the SLO with no operational consequence, or treating it as a target to minimize, when in fact spending it is healthy and a fully unspent budget can signal you are over-investing in reliability. Another error is omitting the policy that gives it teeth.

LIKELY FOLLOW-UPS What goes into an error budget policy? What is burn rate and how does it drive alerting? Who has authority to enforce a freeze? How do you treat planned maintenance against the budget?

ONE CONCRETE EXAMPLE A service has a 99.9 percent availability SLO over thirty days. That permits roughly 0.1 percent of requests to fail, equivalent to about forty-three minutes of full downtime in the month. A messy deploy early in the month consumes most of that allowance through a partial outage. With the budget nearly gone, the agreed policy halts the next risky feature launch and the team spends the remaining weeks improving rollback safety and test coverage. Next month the budget resets, velocity resumes, and the data, not opinion, drove the prioritization.