I would like to suggest and feature for future release and believe that it will be more efficient feature
For upgrade process we are adding new worker node to system and currently it is possible to add more worker nodes or remove from the cluster.
And I haven't seen any documentation about to add or remove Master nodes after suite has been deployed
I am suggesting that because lastly docker has been published that they will deprecated the thinpool feature for future release.
This means that (most porbably) MicroFocus will not use thinool configuration for deployment of SMAX and SMA-SM. (Also MF recommend to use overlay2 configuration for SMAX 2020.02 version currently)
When we consider the currently installed and used thinpool setup on most of customer for SMAX or SMA-SM, they will struggle the keep master nodes with thinpool and only can configure new worker with overlay2 driver during the upgrade process when we intented to add more worker. Then we can remove older worker
So my suggestion is that add new master node and remove the older one, while system is up. This feature could've solve our most of upgrade issues
Because it will install and configure automatically like adding new worker, with no downtime
- We can always change new the new servers with older one, for example; if system resources is not enough, we can add new master then remove older one
- No more get stuck about storage driver configuration for future release
- adding / removing master nodes also allow us to less effort to migrate master servers from older one to newer one