I have a very sporadic issue seen on random desktops. I am in a View 5.0.0 environment.
Occasionally, I have end users indicated that they get the infamous “desktop not available” on their persistent desktop that we are offering up. So I go through my typically test.
Occasionally, but very infrequently, I have seen were the View Agent has simply stopped on the problematic VM. Once I started the View Agent, the end user connected fine. But that is far and few between.
To resolve this issue, a simple reboot clears things up. I have seen this issue for many, many months now. I really need to sink my teeth more into why this is occurring.
Any troubleshooting suggestions I should take?
Thanks in advance...
There is no indication of what the VM was doing. The pattern shows the end user trying to access his VM. It does not kick the end user off per say.
The storage is fine. However, I do have a case open with our SAN storage folks (HP) for a Disk spike that I see every other hour or so. The spike last for a mer 20 to 30 seconds or so. This spike is also see on every ESX (16) server in our VDI farm.
However, I don't believe this is the root of the VDI desktop issue I am describing in this post.
The inability to utilize uptime or connect via services.msc scream some type of guest OS issue to me. Can you login from the vSphere client console when the problem is happening? Can you RDP into the machine when it's happening? That would futher exlude the problem being the View agent and would give you an idea where to focus.
When I access the affected VM from the VI client, the login interface does not come up once I send a C+A+D to the desktop. So when the login credential fields do not appear, I know this VM is truly in limbo per say. I can not RDP to this VM either. So yes, it is a OS thing all the way. The View Agent is simply not running on the VM since it is having chronic issues. The VM will stay this way until it is rebooted. Once it is rebooted, the issue is gone.