Cron reliability checklist after upgrade: verify no skip/duplicate runs

• Cron jobs already exist and you can inspect run history.
• You can create and remove a temporary canary job in non-production channel.

1. Record baseline with cron list/runs for one critical job before upgrade.
2. After upgrade, trigger one immediate force run and confirm it starts/finishes exactly once.
3. Add a 5-minute canary cron and observe at least 3 cycles for skip/duplicate behavior.
4. Check concurrent trigger window by scheduling two jobs at same minute and verify both run independently.

• Clock skew between host and container can still cause drift; keep NTP healthy（需验证）.
• Force runs validate execution path but not long-term load behavior; keep canary monitoring for 24h when possible.

This tip is aggregated from community/public sources and preserved with attribution.