vSAN1

 View Only
  • 1.  Reboot Node when rebalancing is in progress

    Posted Sep 06, 2023 11:01 PM

    Hello,

    I am extending a cluster with one disk group per node. Since the cache device does not seem to be hot plugged, a reboot is necessary on each node.

    Is it better to wait for the cluster to rebalance each time after rebooting a node and adding the disk group?

    Or could I still put the next node in maintenance mode (Ensure accessibility)  during the rebalance and restart without causing problems?



  • 2.  RE: Reboot Node when rebalancing is in progress

    Posted Sep 07, 2023 06:31 AM

    Hi, the best plan would be to power off each host, for the new disks, then repeat for all other hosts. That will result in all new disks being fitted and online in all hosts.

    You can then select to create the new disk groups on each host at the same time and vSAN will then be able to complete just one rebalance task as all new DGs will be created at the same time and no reboots needed.



  • 3.  RE: Reboot Node when rebalancing is in progress

    Posted Sep 07, 2023 08:55 AM

    Hi,

    thanks for that hint. I will do it that way.



  • 4.  RE: Reboot Node when rebalancing is in progress

    Posted Sep 07, 2023 08:54 AM

     Rebalance resyncs place new additional replacement components (e.g. in addition to the data-replicas), once these have finished being written in the new location, the components they are replacing get deleted.

     

    Thus the data retains it's redundancy during this timeframe and MM with EA shouldn't cause any impact (or at least not any more than it would in that cluster where rebalance resync is not ongoing).

     

    That being said, I like  idea of just doing all hardware changes then claiming all the disks after all of them are complete as this might reduce overall amount of data that needs to be moved (e.g. avoid data being moved more than once) so would be advising going with that option..