Why can a scheduled job run more than once in a cluster?
Strong answer
Each application instance may trigger the schedule independently. Use a distributed lock, a single worker, or a coordination mechanism, and make the job safe if duplicate execution still occurs.
