What is Service Level Objective?
Also called SLO, SLI, error budget.
A service level objective is a target value for a measured indicator of service quality over a defined period, such as a percentage of requests served successfully within a latency threshold each month. The indicator itself is the service level indicator, and the gap between the target and perfection is the error budget. Objectives convert vague reliability goals into numbers that guide decisions.
Three related terms are often conflated. An indicator is the measurement, such as the proportion of requests completing under two seconds. An objective is the internal target for that measurement, such as ninety nine percent over thirty days. An agreement is an external commitment to customers, usually with financial consequences and set well below the internal objective so there is room to miss the target without breaching a contract.
The error budget is the practical mechanism. An objective of ninety nine point nine percent over thirty days permits roughly forty three minutes of failure. That allowance is a resource that can be spent deliberately on risky deploys, migrations, and experiments. When the budget is healthy, teams ship faster. When it is nearly exhausted, the same framework justifies slowing down without anyone having to argue from intuition.
Choosing indicators is where objectives succeed or fail. The indicator must reflect what users actually experience, measured as close to them as practical. Server side availability can look excellent while users see failures caused by a broken client path or a slow dependency outside the measured boundary. An objective built on a comfortable but unrepresentative indicator provides false reassurance.
Targets should be set below perfection on purpose. Each additional nine costs disproportionately more, and beyond a point users cannot perceive the difference because their own networks and devices fail more often than the service does. A realistic objective, consistently met and openly reported, is more useful than an aspirational number that is quietly missed every month.
Key points
- Indicator measures, objective targets it, agreement commits externally.
- Error budget is the allowed shortfall in a period.
- Budget state guides how much release risk is acceptable.
- Measure as close to real user experience as possible.
- Targets are set below perfection deliberately.
In practice
A team sets an objective that ninety nine point five percent of agent task submissions are accepted and started within ten seconds, measured monthly. That leaves roughly three and a half hours of budget. A migration in week two consumes ninety minutes of it. With the budget half spent by mid month, the team defers a second risky change to the following period, using the number rather than debate to make the call.