SLA, SLO and SLI compared
In the context of service reliability and performance management, SLA (Service Level Agreement), SLO (Service Level Objective), and SLI (Service Level Indicator) are key concepts used to define, measure, and enforce service quality. While they are closely related, each serves a distinct role in ensuring reliable and predictable service delivery.
1. Service Level Indicator (SLI)
An SLI is a quantitative measure of a system’s performance and reliability. It represents specific metrics that track service health, such as latency, availability, error rate, or throughput. These indicators provide insight into how well a service is performing from an operational perspective.
For example:
- Availability SLI: The percentage of successful requests over a given period.
- Latency SLI: The proportion of requests served within a specified response time threshold.
SLIs are the foundation upon which objectives and agreements are built.
2. Service Level Objective (SLO)
An SLO is a target value or range for an SLI, defining an acceptable level of performance. It acts as an internal reliability benchmark for teams to maintain service quality before breaching contractual obligations.
For example:
- Availability SLO: 99.9% uptime over a month.
- Latency SLO: 95% of requests must be served within 200ms.
SLOs help guide engineering efforts, balancing reliability with operational costs and user expectations.
3. Service Level Agreement (SLA)
An SLA is a formal contract between a service provider and its customers, specifying the promised level of service. It typically includes SLOs, along with consequences (such as refunds or credits) if the provider fails to meet them.
For example:
- Availability SLA: 99.9% uptime, with a penalty if downtime exceeds 43.8 minutes per month.
- Support SLA: A guaranteed response time of under 4 hours for critical incidents.
SLAs are legally binding, whereas SLOs are internal goals.
Comparison and Relationship
Concept | Purpose | Scope | Example |
---|---|---|---|
SLI (Indicator) | Measures service performance | Low-level metric | 99.95% successful requests |
SLO (Objective) | Sets internal performance targets | Engineering goal | 99.9% uptime target |
SLA (Agreement) | Defines contractual obligations | Customer-facing | 99.9% uptime or financial penalty |
Key Insights:
- SLIs are the raw data points, SLOs set acceptable limits on them, and SLAs are enforceable promises based on those limits.
- SLOs are stricter than SLAs to provide a buffer before contractual breaches occur.
- While SLAs focus on external commitments, SLOs help maintain reliability internally, preventing SLA violations.