What is the best way to force a failover from the primary Scheduler to the shadow? The typical method of stopping the Scheduler service via the 'unisrvcntr' command on UNIX or the Windows Service Manager does not cause a failover.
Use the sendevent command to post a 'STOP_DEMON' event with the following syntax...
sendevent -E STOP_DEMON -v FAILOVER
Sending the STOP_DEMON event in this manner will stop the primary Scheduler and tell the shadow Scheduler to take over processing.
If the PrimaryFailbackMode in the instance configuration is set to '1', this same command can be used to failback the shadow to the primary once the primary is restarted. In this scenario, the shadow restarts itself and goes back to shadow mode rather than shutting down completely.