One short definition per term. Link to any of them with its anchor, for example /glossary#burn-rate.
| Term | Definition |
|---|---|
| Availability | The share of time, or of requests, for which a service works as intended. Usually expressed as a percentage such as 99.9%, and often used as an SLI. |
| Burn rate | How fast an SLO is using up its error budget, relative to the pace that would use it up exactly by the end of the window. A burn rate of 1 spends the whole budget in the window; 10 spends it ten times faster. Read more |
| Change failure rate | The share of deployments or changes that cause a degradation or incident and need a fix, rollback or patch. |
| Context graph | A live, connected model of an IT estate that links SLOs, services, infrastructure, changes, code and ITSM records, so a symptom can be traced to its cause. Read more |
| Error budget | The amount of failure an SLO allows over its window: 100% minus the target. For 99.9% over 30 days, about 43 minutes of full downtime. Read more |
| Error budget policy | A written agreement on what a team does as error budget runs low, such as slowing releases or freezing them. Read more |
| Incident | An unplanned interruption or degradation of a service that needs a response. In ITSM, an incident record tracks it from detection to resolution. |
| MTTR | Mean time to recovery (or repair): the average time from the start of an incident to the service working again. |
| Multi-window alerting | Checking burn rate over a long and a short window together so an alert fires only when a burn is both significant and still happening. Read more |
| Observability | The ability to understand what a system is doing from the telemetry it emits: metrics, logs and traces. |
| OpenSLO | An open, vendor-neutral specification for declaring SLOs, SLIs and related objects as YAML. |
| Pre-breach SLO management | Keeping SLOs met by acting on error budget burn rate before the SLO breaks: detection, correlation, approved remediation and a code fix in one loop. Read more |
| Problem management | The ITSM practice of finding and removing the underlying cause of one or more incidents so they do not recur. |
| Rollback | Returning a service to a previous known-good version or configuration to stop a degradation caused by a change. |
| SLA | Service level agreement: an external, usually contractual promise about service performance, often with penalties or credits if missed. Read more |
| SLI | Service level indicator: the measurement an SLO is judged on, such as the share of requests that succeed in under 300 ms. Read more |
| SLO | Service level objective: an internal target for how reliably a service should perform over a window, such as 99.9% over 30 days. Read more |
| SRE | Site reliability engineering: the discipline of running production systems with engineering practices, using SLOs and error budgets to balance reliability and change. |
| The nines | Shorthand for availability targets by the number of nines: three nines is 99.9%, four nines is 99.99%. Each extra nine allows ten times less downtime. |
| Toil | Manual, repetitive operational work that scales with the size of a service and adds no lasting value. A target for automation. |