ESXi 6.0.0 9313334 - Windows server 2012R2 with VMXNET2 NIC random stuck

[scoulon]

     Dear all

Since installation of the August updates (ESXi600-201808001), a windows server 2012R2 VM (using VMXET3 NIC) randomly stucks : all RDP session are disconnected, we are not able to log on locally because the VMWAre console doesn't respond to CTRL+ALT+SUPPR. If the console is already logged for a local administrator with the task manager, the system seems freezed.

In vCenter, we can see a high level of vCPU for this VM when it freeze.

All other VM continue to run without problems.

Problem has not been solved after installation of September updates (ESXi600-201809001).

We have tried to upgrade client vm-tools to version 10.3.2 for this VM : but the problem is still there.

At the moment, he only way to get the system back is to force the VM to power off and start it fresh new.

Do you have some ideas ?

     Best regards

[RajeevVCP4]

Is this issue only windows 2012 ?

Did you apply any patch on windows VM ?

Did you upgrade virtual hardware ?

attach vmware.log file , with time stamp when issue was started.

[scoulon]

You'll find the vmware.log file in attachment. It correspond to the last freeze we have for this Windows 2012R2 VM.

It was stuck during the mid-day pause. It seems to correspond to event :

     2018-09-24T10:19:08.203Z| vmx| I125: GuestRpcSendTimedOut: message to toolbox-dnd timed out.

VM has hardware version 11, with 4 vCPU / 16 GB of RAM.

All available windows updates patches are installed in the VM.

Only this windows server 2012R2 has the problem, but it's the only windows 2012R2 VM we have on the server. All other VM have different version or OS (linux debian, linux centos....).