Hi,
I’ve noticed that if I change the NVIDIA vGPU Profile for example from M6-8a to M6-2a the MCS machines won’t boot up.
They are ending in a boot loop with trying to "getting devices ready".
Is this behavior expected, or is something wrong on my machines?
Evironment:
VMware ESX 6.5
Win Server 2016, XenApp 7.18
NVIDIA M6, Host Driver 390.57, Guest Driver 391.58
Thanks a lot.
Hi
The Operating System will see the change in vGPU profile as a change in GPU, so it may be trying to install what it thinks is new hardware. On your Master VM, change the vGPU profile to the required one, start it up, let the Operating System detect it and make sure it’s in use in Device Manager. A reboot will be required. Then shutdown, take a snapshot and push out the new image through MCS and your VMs should then be ok as long as they are running the same vGPU profile as the one on your Master VM.
Just to add … You don’t want to use a 2A profile for XenApp, especially on an M6. 8A is the recommended profile for XenApp ;-)
Regards
Ben
I agree with using a much larger profile for your XenApp instance. Make sure to also allocate a reasonable amount of memory and number of VGPUs to your XenApp instance. Monitor it to make sure your resources are then picked to be about right and tweak them, if necessary to better optimize it.