I haven’t found in the Guide the recommended procedure for rebooting the master node. When it had fewer CryoSPARC instances, I could wait until none had running or pending jobs to reboot. Now that no longer occurs. There are a mix of v5.0.7 and v4.7.1 instances. I might guess this procedure:
- for each instance, turn on maintenance mode
- wait for all running jobs to complete
- reboot master node
- after reboot, start each instance
Are there implications or risks of this method that I’m overlooking? Would anything different be recommended? Should the instances also be stopped before the reboot? Does starting an instance automatically take it out of maintenance?