1 person found this helpful
Check Alert log of Database Instance on that node.
By default if Oracle Database Instance already is started in one node, CRS automaticaly start Database Instance automatically on others nodes.
But if you issue "srvctl stop instance" before shutdown of node the state of this resource will be "shutdown" (i.e stay down) CRS database resource have by default the attribute AUTO_START=restore, which means Oracle CRSD will remember last state of that resource.
In this case you must manually issuing "srvctl start instance" after startup of clusterware, but if database instance was running and you issued "crctl stop crs", crsd must start database automatically at clusterware start.