Observe and troubleshoot

Start with resource, deployment, runtime, proxy, and access route status before assuming the whole deployment failed.

Runtime logs come from application stdout/stderr and help identify startup, port, configuration, and runtime failures.

The health summary combines deployment, runtime, health policy, proxy, and public access observations.

Diagnostic summaries should include stable IDs, status, error codes, and safe context while masking secrets.

Prefer retryable operations first. Change server, credential, proxy, or domain configuration only when the status explains why.