This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
Has anyone experienced an issue with Azure Virtual Desktop VMs where a VM or two in the pool will, randomly, stop accepting connections. No Users can login to the pool until the affected VM is put into drain mode at which point users can login. Trying to just RDP directly into the affected VM gets stuck at "securing connection" and just sits there until it times out. If you reboot the VM, you can get in and it works fine.
The VMs have a lot of software on them and are running NVIDIA GPUs (they are NV series).
I can telnet to the affected VMs on 3389, just cant connect via RDP after the securing connection status just sits there.
The logs have some errors which I will post below, but I thought Remote FX was removed by Microsoft a while ago. We even resized the VMs from NV12 to NV24.
These are 21H1 VMs that had been running fine for about a month and all the sudden are experiencing this issue. We are trying to talk to Microsoft but wanted to ask here as well.
Seeing these errors from Azure Insights (cant connect to the affected machines to get the errors directly)
Attempt to send disconnect (138) message to Windows video subsystem failed. The relevant status code was 0xffffffff80070102.
An error occurred when transitioning from LoggedOn in response to EvDisconnected. (ErrorCode 0xffffffff80070102)
An Azure service that is used to provision Windows and Linux virtual machines.
A Microsoft desktop and app virtualization service that runs on Azure. Previously known as Windows Virtual Desktop.
We are using the NVIDIA driver that installed with the extension when we selected it.
Is that correct?
Plus not happening to all VMs, just random ones at times. Most have had it happen but they all havent at once.
Running 461.33 right now, I just checked.
