15.8 Evidence Collection

Go To Top

Return to Logging and Observability

Evidence collection preserves the results of setup, generation, build, run, validation, and acceptance activities. Evidence allows the team to review what happened after the system stops and to compare results across runs.

The team should collect Evidence for:

1. Prerequisite checks.
2. Tool baseline verification.
3. DDS/IDL type generation.
4. Build results.
5. Container image creation.
6. Node startup and shutdown.
7. Control Plane status reporting.
8. Control Plane command handling.
9. Data Plane message flow, where applicable.
10. Test and validation results.
11. Acceptance criteria.

Evidence may include:

1. Command output.
2. Script logs.
3. Node logs.
4. Container logs.
5. Generated Artifact manifests.
6. Captured DDS messages.
7. Test reports.
8. Screenshots, where useful.
9. Reviewer notes.
10. Checklist instances.

Evidence should live in an approved Evidence location, such as:

evidence/

The team should organize Evidence by baseline, run, date, or acceptance activity. A representative structure is:

evidence/
    phase0/
        run-001/
        run-002/
        acceptance/

Evidence should identify the baseline, run date, tool context, and relevant scripts or Nodes. The team should not rely on undocumented terminal history or personal notes as Evidence of acceptance.

The practical rule is simple: a reviewer should be able to inspect the Evidence and understand what the team ran, what the system reported, what passed, what failed, and what remains unresolved.