RED Method
The RED method analyzes three metrics per service: Rate (request rate), Errors (error rate), and Duration (latency), monitoring service health.
Purpose
Systematically monitor microservice request health, driving alerts and SLOs.
Mechanism
Rate: requests per second. Errors: failed requests per second or ratio. Duration: latency distribution (p50/p95/p99). Per-service dashboards show R/E/D. Complements USE: USE watches resources, RED watches services.
Use cases
- Service monitoring
- Alerting
- SLO