sre-practices
Installation
SKILL.md
SRE Practices
Apply Site Reliability Engineering principles.
Service Level Concepts
| Term | Definition | Example |
|---|---|---|
| SLI | Measured metric | Request latency p95 < 500ms |
| SLO | Target threshold for SLI | 99.9% of requests < 500ms |
| SLA | Contractual commitment (usually looser than SLO) | 99.5% uptime |
Choosing SLOs
- Pick metrics users actually care about
- Start with availability + latency + durability
- Tighten SLOs over time as reliability improves
- Don't over-constrain (cost/effort vs. benefit)