[Feature request] HPE Morpheus Enterprise with 3-nodeHA

[kountory]

I have an existing case number 08047302

We would like to request a modification to the software to enable the registration of an HPE Morpheus 3-node HA configuration.

When adding HPE Morpheus Enterprise with a 3-node HA server configuration, an error occurred.
The message indicated:
“Failed: Failed to scan VMs: Error occurred while deserializing the response.”
Therefore, we created a support ticket. The support engineer informed us that the Veeam Plug-in for HPE Morpheus VM Essentials does not currently support an HA configuration for the Morpheus Manager.

Veeam Backup & Replication 13.0.1
HPE Morpheus Enterprise 8.1.0

[cody.ault]

Hey, can you tell me a bit more about the configuration?
Is your configuration something like this? > https://support.hpe.com/hpesc/public/do ... E0827.html
When you added the Morpheus Enterprise environment, did you use a load balancer IP? We didn't test this configuration, so I'm not sure if it's related to the error you saw or if there's a different problem and the HA configuration is just distracting from the actual issue.

When I looked at the case, it looked like there was a note about an error with the VM(s) having null for one of the parameters we expected to have a value related to storage. Are you using a GFS2 or NFS data store for you VMs? I'm asking the support team to send over the logs so we can review. Any information about how the management servers are setup will help reproduce the problem.

[kountory]

Thank you for your reply.

It follows the structure of the URL you provided.
I tried registering both the load balancer VIP and the individual IP address, but encountered the same error.

The VMs are using GFS2 datastore.

Thank you for analyzing the logs. We appreciate your continued support.
Please let us know if there is any additional information required.

[cody.ault]

Based on the errors from the logs, the issue seems to be related to the VMs we're unable to scan due to the null value and I don't think it's related to the Morpheus Enterprise HA configuration. We'll try to validate the configuration for the next release. To be able to continue adding this particular environment, you might need to work with HPE to address the missing values that are causing the "failed to deserialize" message.

[cody.ault]

A quick update, I setup a 3 node Morpheus Enterprise environment in my lab and it seemed like everything worked fine. We'll look to validate the configuration officially and note that it's supported in a future release, but it shouldn't be a blocker as far as I can tell.

[fukudas]

Excuse me for butting in.
I’ve received an update stating that everything is working without issues in a 3-node Morpheus Enterprise lab environment.
I’d like to verify this in my own environment, so I’d appreciate it if you could provide me with the lab environment details.

Could you please provide the following information regarding the environment where the backup succeeded in a 3-node HA configuration?
- Morpheus version,
- Host agent version,
- morpheus-ws-node package version,
- Veeam version,

[cody.ault]

3 Ubuntu 24.04 VMs
- Morpheus version > 8.1.2
A mysql 8.0 deployment
A HAproxy VM

Host installed on HVM24.04
- Host agent version > Agent Version: 3.1.4
- morpheus-ws-node package version > 3.3.25-1
- Veeam version is the GA version of the Morpheus Plugin (13.1.0.271), the March patch of VBR (13.1.0.2067).

The support case related to the original post in this thread is failing due to specific VMs that have null values in their metadata that shouldn't be allowed to be null. My assumption was that perhaps these VMs were created through libvirt natively and discovered and converted to managed but I don't have that information.