Find the right diagnostic procedure

The operational runbook is Troubleshoot by symptom. It separates startup, discovery, request security, persistence/replication, and consumer completion.

1. Start with first diagnosis and record one failed operation.
2. Use Admin observations to compare routes, Broker state, and consumer progress.
3. Use monitoring to correlate controlled logs, metrics, and traces.
4. Preserve data before a recovery decision; use maintenance for a planned restart.

Rust processes do not expose JVM heap/thread-dump interfaces. Use the actual Rust process, supported service log filters, and platform observations. Collect stable public errors and non-secret context; keep credentials and message bodies out of reports.