15.4 Node Logs

Go To Top

Return to Logging and Observability

Each Node should produce logs that explain its Lifecycle and runtime behavior. Node logs should support diagnosis, review, and acceptance of Evidence.

Node logs should record:

1. Node startup.
2. Configuration source.
3. Node Identity and role.
4. DDS participant initialization.
5. Publisher and subscriber initialization.
6. Topic participation.
7. Lifecycle state changes.
8. Status publication.
9. Command receipt and command result, where applicable.
10. Data Plane publication or subscription events, where applicable.
11. Warnings and degraded conditions.
12. Recoverable and non-recoverable errors.
13. Controlled shutdown.
14. Exit result.

Node logs should align with Control Plane status messages. If a Node publishes Running status, its logs should show the successful startup path that led to Running. If a Node publishes Failed status, its logs should identify the failure condition. If logs and status messages disagree, the team should treat that disagreement as a defect or open issue.

Node logs should use the file header and inline documentation conventions defined in Section 7.10 where logging functions, logging wrappers, or logging configuration files require explanation.