SLI (service level indicator)
A service level indicator (SLI) is a number that measures one aspect of the service users receive, such as availability or latency. It is usually written as good events divided by valid events, expressed as a percentage. A service level objective (SLO) then sets a target for that SLI over a time window.
How it works
Pick an event users care about, decide what counts as good, and count both. Common SLIs:
Availability: requests that did not return a 5xx error.
Latency: requests served faster than a threshold, such as 300 ms.
Freshness: data updated within the last 10 minutes.
If 100,000 checkout requests arrive in a day and 99,870 finish in under 300 ms, the latency SLI is 99.87%. Measure as close to the user as you can, for example at the load balancer rather than inside one pod.
Why it matters in production
An SLI turns "is the service OK?" into a number everyone reads the same way. Bad SLIs produce bad decisions: an average latency SLI can look flat while the p99 tail hurts users. Good SLIs make error budget math and alerting trustworthy.
See also
On the One2N blog
Further reading
