etcd-io/etcd
How to monitor
etcd is a service. It runs forever, and the people who run it want to see what it's doing.
Pages
- Logging — zap-based structured logs.
- Metrics — Prometheus metrics exposed at
/metrics. - Tracing — per-request
traceutiland OpenTelemetry.
Quick start
| Need | Endpoint / Path |
|---|---|
| Prometheus metrics | :2379/metrics |
| Health (HTTP) | :2379/health |
| Health (gRPC) | grpc.health.v1.Health/Check |
| Server version | :2379/version |
pprof (when --enable-pprof) |
:2379/debug/pprof/* |
| OpenTelemetry traces | --experimental-enable-distributed-tracing=true --experimental-distributed-tracing-address=otel:4317 |
The Grafana mixin in contrib/mixin/ ships dashboards and alerts that use these endpoints.
Built by Factory AutoWiki from public repository content. It is a generated preview for codebase exploration, not source-maintained documentation.