Hello, there! Recently we came across a very very strange behavior when trying to remove AOS instances from the cluster config of AX 2012 R3... Here's the story...
We have a PROD instance with 5 AOS instances - 4 AOS in a cluster and one out of the cluster that serves as a batch server. All works as a charm. Then we take a backup of the PROD transnational DB and restore it over the TEST DB. The TEST environment is a sandbox that consists only of one AOS.
Now, after the restore, we open the Cluster configurationof the refreshed TEST environment and we see the 5 AOS instances from PROD along with the currently active TEST AOS instances.
![]()
We need to remove those 5 PROD AOS records from the cluster configuration of TEST. If we try to do so, we get the following error:
![]()
As per my understanding, if we want to remove those application servers from the cluster config, they need to be drained from client connections and stopped, no matter they are actually serving another instance? OK, here comes the strange part... After 2 attempts to remove them without stopping anything, I got that same mistake. THEN, on my 3rd attempt, I actually managed to remove all 5 PROD AOS instances, as well as the whole cluster from the config:
![]()
Any idea why and how this actually happens? First I get the error, then nothing changed at all, only 10 minutes later I manage to remove all app servers with no issue... Is there any mechanism behind this?