rename managed server inside a failovercluster not working with KB1905

[mairec_it]

Hello guys,

I want to get rid of some last NTLMv2 communications. To get kereberos working I need the FQDN everywhere. Currently my problem is that I can not rename nodes inside a hyper-v-cluster. I used KB1905 to rename the cluster under Managed Servers - that worked well.

Now I have my cluster.fqdn but the nodes inside are without .fqdn just node1, node2. When I use powershell to rename node1,node2 to node1.fqdn and node2.fqdn gets overwriten by the next rescan.

Anyone got a solution for this?

I already tried to remove a node within the failover manager and add it back with fqdn, but this does not change anything.. Currently the support is not an big help to get this sorted (Case # 07324564)

[david.domask]

Hi mairec_it, welcome to the forums.

KB1905 uses unsupported methods and the KB needs to be updated on its limitations -- the node name changes are likely happening because KB1905's method doesn't "touch" all the places for a cluster.

I'm afraid that removing and re-adding the cluster will be the only supported way of accomplishing this, KB1905 likely will not work for such clusters.

[mairec_it]

I now have tried to remove the cluster. After removing, I added it back with fqdn, still when in "Review"-Step both node1, node2 server are displayed without fqdn part..

[david.domask]

Hi mairec_it,

Sorry to hear that there are continued challenges -- please update your case with the above information and continue working with Support on the situation, as troubleshooting will not be possible over the forums. May I ask, had you restored from a Configuration Backup after performing the steps in KB 1905? My guess is that right now there is an inconsistency in the configuration database because of the changes from KB 1905 and likely it would be best to revert prior to the changes from KB1905 -- the Configuration Restore will initiate a rescan of the repositories, replicas (if used), and tapes (if used) to ensure that backups made after the Configuration Backup are found and updated in the Configuration Database, but likely it's best to get to a "clean slate" prior to the changes made with KB 1905.

[mairec_it]

I tried some other things but cant get it to work. When removing a node from the hyper-v-cluster and adding the node back with fqdn. The summary show that adding was successfull with FQDN but under nodes it is again listed without fqdn
So there is currently no way in getting this sorted?

[mairec_it]

Hi,

someone got a solution for this? I'm still struggling with this topic to finally fully disable ntlm..

[david.domask]

HI mairec-it,

Can you clarify the current issue? Are you having troubles with backups after ensuring the requirements from the user guide page for Kerberos support?

That's the key point, the backing configuration. KB1905 should not be needed here.

[mairec_it]

No, kerberos is not an option as the hyper-v-cluster with containing Hosts do not speak in FQDN. Thats why veeam uses still NTLM to connect to them. I had a support-case for this open, buy no solved the issue with me.

The cluster-object under managed servers are in FQDN but the Hosts itself are not, like mentioned:
https://imgur.com/ei6p5PD

How can I achive this or am I misunderstanding something to get rid of NTLM?