sbcsu asked for sta-hista:search; which gives you the last 2 hours of messages, but you did sta-hista; which only lists the alarms table. Alarms table shows a CC is restarting but you'd already said that.
Cannot tell from your message whether the CCB VM is restarting, or the whole platform for node 2 is restarting. If you login to the linux, "uptime" will tell you. If the uptime is a long time but you know it restarted a few hours ago, then it looks like you have an RMX problem which is maybe a little less serious than a HW problem on the processor.
But in either case, remedial action is the same. If corosync, atlans connections are all OK and processor A can reach processor B on both networks, then node 2 should be reinstalled.