|
In your case above, if you split up 200 instances into 4 groups, you will lose 1/4 of the capacity during the upgrade. Not the worst thing to do, but it doesn't pass the smell test to me. These groups seem artificial and like a workaround to me.
My use case is cleaner by far, as you only lose one node's worth of capacity during the upgrade,
But back to your point, it is indeed true that with groups, you can "get around" the deployment issue with WFLY only (as far as I know) This by no means makes my use case less valid.
|