question

MarekGodlewski-9034 avatar image
0 Votes"
MarekGodlewski-9034 asked MarekGodlewski-9034 answered

VMs restart after a node came back from pause

Hi,

We've a HV cluster based on Windows 2012 R2 consisting of 4 nodes connected via FC to common storage (NetApp). From couple of months we are getting strange behavior. When we put a node into to the pause al machines are moved from this node correctly. Then we are making any maintenance like OS update etc. Then we restarting this paused node - after it get back to live we are unpause it. And then strange things happens - all the vms are returning back to the node but all of them resets dirty :(. We were checking all of the logs and have no clue what can be wrong with our configuration. During unpause the communication with the cluster is ok also the storage is available. Maybe some of You have any clue or suggestion ?

best regards

Mark

windows-server-hyper-vwindows-server-clustering
5 |1600 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.

didier3001 avatar image
0 Votes"
didier3001 answered

Hi @MarekGodlewski-9034

Is there any error or even "Information" in the Windows event log on any of the Hyper-v node that could be relevant and could help us identify the problem?

Can you also post the output of the following commands:

  1. Get-ClusterResource "Virtual Machine"
    => Replace "Virtual Machine" with the name of one of the VM you know 100% had the behavior your described in your original post.

  2. Get-ClusterResource -Name "YourClusterName" | Get-ClusterParameter
    => Replace "YourClusterName" with the name of your cluster

Regards,
Didier3001


5 |1600 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.

MarekGodlewski-9034 avatar image
0 Votes"
MarekGodlewski-9034 answered MarekGodlewski-9034 commented

Hi @didier3001

ThankYou for the reply. Here is what I noted for the VM named WEDWARESET03 during it restart at 5:21 PM on 08.08.20. I'm attaching part of the VM system log from that time and also part of the Cluster log from HV (WEDWARHV06) host on which this VM resides. Also attaching results of the ps commands You mentioned. Second command should be:
Get-ClusterResource -Cluster "YourClusterName" | Get-ClusterParameter ?

best regards

Marek Godlewski16564-hv-wedwarhv06-clusterlog-1715-1730-080820.txt16593-wedwarhvc01-cluster-parameter.txt16565-wedwareset03-information.txt16556-vm-wedwareset03-system-log.txt



· 2
5 |1600 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.

I don't see anything obvious in the logs. OfflineAction is configured to the default value of 1 which is good.

Can you run the failover cluster validation tool and make sure you select the Hyper-V component and then, send us the output in HTML format?
https://docs.microsoft.com/en-us/previous-versions/windows/it-pro/windows-server-2012-R2-and-2012/jj134244(v=ws.11)#to-run-the-validate-a-configuration-wizard

Thank you,
Didier3001

0 Votes 0 ·

Yes I fighting with this problem from couple of weeks and also did not find any obvious errors. I will make cluster validation tests today during evening hours but probably without storage tests as it will temporarily stop the cluster and I need much more approvals from all business divisions.

best regards

Marek Godlewski

0 Votes 0 ·
MarekGodlewski-9034 avatar image
0 Votes"
MarekGodlewski-9034 answered

Hi - In the meantime we noticed those entries in cluster log (screenshots attached) - maybe this can tell some about the problem ?

best regards

marek Godlewski17727-image002.png17728-image001.png



image002.png (116.8 KiB)
image001.png (105.3 KiB)
5 |1600 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.