SLA (service level agreement)
A service level agreement (SLA) is a contract between a provider and its customers that promises a level of service and states what happens if the provider misses it. The promise is usually an availability or response time target. The consequence is usually financial, such as service credits or a refund.
How it works
An SLA is built on a service level indicator (SLI) and a target, plus a penalty. For example, a contract might promise 99.9% monthly availability, which allows 43.2 minutes of downtime in a 30 day month, and credit 10% of the bill below that.
Teams set their internal service level objective (SLO) tighter than the SLA, for example 99.95% (21.6 minutes). The gap is a safety margin: the team reacts when the SLO is at risk, long before the contract is breached.
Why it matters in production
An SLA turns reliability into a business commitment that sales, legal and finance care about. Promising more nines than the architecture can deliver creates real liabilities. Engineers should check every SLA number against measured SLIs before anyone signs it.
See also
On the One2N blog
Further reading
