Hello everyone,
I have been searching for the proper way to recycle each of the nodes in my 3 node, HA vROPS 6.2 Cluster so that data is not lost during the node reboots.
Here is the setup:
Master Node
Replica Node
Data Node
My question is how do I take a single node out of service correctly (say for a config change/reboot?) so that running adapter instances move to another node and no data is lost during that timeframe?
The process I followed was to take the data node out of the cluster first, wait for that the say offline, then reboot it, however I noticed that the adapters that were running on it didn't move and data was lost.
I proceeded to see if this mechanism might only work for the master/replica set (which wouldn't make sense, but I tried anyway) which had the same result, the replica node was taken out of service and rebooted, but the adapter instances on it went to not collecting and I'm getting No Data for any of those vCenter objects and if I got to click on them, thinking maybe the cluster mgmt screen was behind, I get this message when the object loads up "No such resources in shard mapping"
Can anyone tell me how I am supposed to recycle a node without data loss? Do i have to move the adapter instances manually for each node I want to reboot BEFORE I offline it from the cluster and reboot? Do i just reboot them to use a "failover" mechanism so that the adapters fail over automatically?
I can't find any documentation and calling VMware support is painful for questions like these. If anyone can enlighten me, I would greatly appreciate it! Thanks!
