---
title: "Service Level Indicator (SLI) — monitoring glossary"
description: "A Service Level Indicator (SLI) is the actual measurement of one aspect of service — usually availability, latency, error rate, or freshness. SLIs are the…"
canonical: "https://lemwatch.com/glossary/sli"
source: "https://lemwatch.com"
---

# Service Level Indicator (SLI)

> A Service Level Indicator (SLI) is the actual measurement of one aspect of service — usually availability, latency, error rate, or freshness. SLIs are the numbers you compare against SLOs to decide whether the service is meeting its reliability target.

## Definition

 Good SLIs are ratio-based: good events divided by total events. 'Percentage of homepage requests returning HTTP 2xx within 400ms' is a strong SLI. 'Average response time' is a weak one — averages hide the tail latency users actually feel.

 Every SLO needs a matching SLI wired into monitoring. Lemwatch surfaces the raw SLI numbers per check (uptime %, P50/P95 latency, error rate) so teams can define SLOs against them directly.

## Why it matters

 Without a measured SLI, an SLO is aspirational. Wire the SLI into the alerting stack before publishing the SLO — otherwise the team can neither prove compliance nor detect breach.


## Related

- [Monitoring glossary](https://lemwatch.com/glossary)
- [Service Level Objective](https://lemwatch.com/glossary/slo)
- [SLA](https://lemwatch.com/glossary/sla)
- [Uptime](https://lemwatch.com/glossary/uptime)
- [Error Budget](https://lemwatch.com/glossary/error-budget)
- [Real User Monitoring](https://lemwatch.com/glossary/rum)
- [All monitoring checks](https://lemwatch.com/features)