Over the weekend, we took a planned power outage in the building that houses our test environments. Due to external factors, we were only able to perform graceful shutdown (using Configuration > Power > Power Off Appliance) on half the appliances. The other half had their power cut while they were running.
When power was restored to the building, the appliances with dirty shutdowns came back without issue. The appliances with clean shutdowns did not come back on, and required physical intervention to restart.
Is this expected behavior? If so, is the recommended procedure for a planned power outage to stop the cluster but let dirty shutdown happen?
A Power Off action switches the appliance off. This state is not changed by a temporary datacenter power loss.
Release : 3.3
Component : PRIVILEGED ACCESS MANAGEMENT
An appliance that was powered off deliberately, will have to be powered on manually. This is the recommended procedure. However, if there is significant overhead involved in getting someone to the hardware appliances to power them on, it is acceptable to leave them powered up, in which case they should come back on automatically after power is restored. The most important step is to turn the cluster off prior to a scheduled power outage. Database replication recovery can be problematic if active nodes were down for an extended period of time.