Observability

Thresh emits signals through logs, metrics, and tracing to support production diagnosis.

Logging

  • HTTP pipeline start/stop events
  • Retry/circuit events (via configured resilience components)
  • Event stream lifecycle transitions

Use structured logging sinks and retain endpoint/path dimensions where possible.

Metrics

The core metrics surface captures request counts, durations, and failure counts. Track trends over time instead of relying on single-point samples.

Tracing

ActivitySource("Thresh") spans can be exported through OpenTelemetry-compatible pipelines.

  • HTTP request rate
  • HTTP error rate
  • HTTP latency distribution (p50/p95)
  • reconnect count/rate
  • event handling throughput (application-level)