Restart Recovery Drill
What happened
tc drill restart-recovery --run-id <run_id> performs an artifact-based recovery rehearsal by checking run artifact presence.
Expected behavior
- Command exits
0. - Drill report is written to
artifacts/runs/<run_id>/drills/restart_recovery.json. - Report includes a checklist summary for required and optional artifacts.
- Report includes: "no process restart performed; this is artifact-based recovery rehearsal".
- If
journal.jsonlexists, a compactfailure_drill_executedevent is appended.
How to inspect artifacts
- Open
artifacts/runs/<run_id>/drills/restart_recovery.json. - Confirm:
outcome=simulated_recovery_oksummary.metadata_present=truesummary.run_spec_present=true- statement explains no real restart was performed.
What a real production system would do
- Coordinate supervised process restart.
- Validate warm-start state and downstream connectivity.
- Run post-restart safety checks before re-enabling trading actions.
What this lab intentionally does not do
- No process orchestration.
- No daemon/supervisor control.
- No background watcher or async recovery loop.