SLIsSLOs

SLIs and SLOs are concepts used in reliability engineering to measure and commit to service performance from the user’s perspective. An SLI, or service level indicator, is a quantitative metric that reflects a dimension of how a service performs for users. Examples include availability (the fraction of successful user requests), latency (response time, such as p95 or p99), error rate (the share of failed requests), and throughput. An SLO, or service level objective, is a target value or range for an SLI over a defined period, such as 99.9% availability per month or a p95 latency under 400 milliseconds. SLOs are the concrete commitments teams make to meet user expectations.

SLIs and SLOs are often used together as the technical foundation of SLAs, which are formal agreements with customers that may include remedies for breaches. The SLI/SLO approach focuses on reliability and user experience, while an SLA may include broader commercial terms.

Setting effective SLOs involves selecting SLIs that truly reflect user impact, choosing targets that balance reliability with development velocity, and defining a measurement window (commonly monthly). An error budget—calculated as 1 minus the SLO (for example, 0.1% allowable failure for a 99.9% SLO)—guides prioritization, incident response, and release planning. Data must be accurate and auditable, with clear definitions for how measurements are collected.

Practically, SLOs guide monitoring, alerting, capacity planning, and post-incident reviews. They should be reviewed regularly and adjusted as user expectations and product requirements evolve. Common pitfalls include selecting impractical or poorly measurable SLIs, misaligning targets with customer needs, and failing to maintain data quality.