Working with a customer on issue remediation, one of the problem the customer had was that the MAM LB on his 3 nodes of cluster was not working.
To describe a little bit the environment, the customer uses in production 2 MPX 8200 in HA with 2 partitions for security purpose.
XMS are declared on DMZ Domain 0 and LAN Domain 10.
So we can easily test independently MDM and MAM LB feature.
So MAM was not working and I need to investigate on this problem.
I will not detail all the steps and checks I made but just let you know what was the issue.
In fact, the customer used the Wizard for the configuration but provided all the commands manually on CLI (to have his own naming convention applied to XMS/NetScaler configuration).
The problem was that on the MAM Service Group, he had the first node with the correct CustomID but the other nodes were added based on the first, just with changing the name and not the ID.
So customer was trying to configure MAM LB with 3 Nodes and same ID defined on the Service Group.
So what happened:
Each node was able to work correctly individually but trying with any of those node with 2 or 3 in the Service Group caused issue with the MAM.
When more than 1 node is configured with same ID communication , issue appears.
If you have any issue with MAM and you configured your NetScaler manually, just check that MAM Service Group has each Nodes of your cluster with different (and correct ) ID .
The ID can be retrieved in the XMS Console, under Troubleshooting and Support > Cluster Information
