Operations

Observability, ClickHouse traces, troubleshooting, health checks, and incident response.

Run and debug the shipped auth + proxy stack — metrics, traces, probes, common failures, and runbooks.

OpenTelemetry, Prometheus, and structured logs.

Async LLM trace writes and org_id filtering.

Common compose-dev and service failures.

Liveness, readiness, and probe contracts.

High-value runbooks for on-call.

Was this page helpful?

Last updated on