ERROR Failed to create WMI invoker for \root\mscluster - Invalid namespace

[kaffeine]

Hello everyone,

we're using Veeam ONE to monitor a few hypervisors which are running Windows Hyper-V Server 2019 (the Hyper-V only SKU) on standalone mode (not clustered).

We updated our Veeam backend to V12 and unfortunately we're dealing again with an old known (and reported) bug: Veeam ONE tries permanently to access an invalid WMI namespace:

Code: Select all

ERROR Failed to create WMI invoker for 10.112.112.103\root\mscluster - System.Management.ManagementException: Invalid namespace 
ERROR [CSVLOAD - 10.112.112.103] Failed to load MSCluster_ClusterSharedVolume System.Management.ManagementException: Invalid namespace 

Our DataProvider_HyperV logs get filled up with this error at a crazy rate: 2 to 4 log lines per hypervisor are generated every minute, and this goes on 24x7.

We were previously using the Windows Hyper-V Server 2012 R2 SKU, and we reported this erratic behaviour with the \root\mscluster namespace back in 2022 (#05769191). Back then the feedback was "fix will come with V12". After updating to V12 we saw it was not the case, and submitted a ticket again (#07290216), cause this time these constant queries to an non-existant WMI namespace where also causing problems with the performance collection.

After a long back and forth with the support, the feedback is once more that there's no fix to this behaviour, and a solution "might" eventually come (or not).

We understand that is an edge case (this SKU has probably a much lower adoption compared to the standard Windows Server SKUs - which do have the \root\mscluster namespace), but it is frustrating that no action at all is taken from the development to minimize this crazy amount of generated error logs. After all, we are using an officially supported OS.

Is anyone else with the Hyper-V only SKUs dealing with the same issue?

Regards

PS - here is a snippet from the DataProvider Logs

[jorgedlcruz]

Hello,
We have reviewed this topic internally, we have it documented, and a proper bug was created. I discussed it internally, and we will try to fix it after v12.2 release, as a hotfix.

That means we should have a fix sometime in early September, perhaps sooner. As soon as I have some more news, will let you know here, but usually Support will contact you with the hotfix.

Thank you for raising the question again, and thank you for the patience.

[kaffeine]

Hello Jorge,

those are great news, thank you for acknowledging the problem and handling it internally - and thanks to Veeam for making this R&D forum available and allowing a direct dialog.

We look forward to the hotfix!

Best regards.