What happens to schedule jobs when node restartThere are only two broad categories of jobs when it comes to node restart/crash behavior because of any reason (like patching or accidental restart because of low heap memory ) Pinned Jobs (sys_trigger.system_id != NULL) All of these jobs are child jobs (e.g.events processor, Flow Engine Event Handler etc.) which gets deleted automatically when the node is marked as offline in sys_cluster_state table. When the node comes back up the child job is auto-created for that node. Unpinned Jobs (sys_trigger.system_id = NULL) These are run-once/async-BRs or repeat jobs (with system_id = <empty>). When the node crashes while the job is Running state, it gets recovered i.e. state changed to Ready by Scheduler. Then it might run on a different node if available otherwise same node if it comes back up sooner. NOTE: Graceful Shutdown feature added from Utah onwards reduces the chances of node shutting down for maintenance/patching while jobs are running on them by waiting for 5 mins before node restart so that if job is quick it can get complete.