Automatic Updates Reboot Doesn't Work in P2Ved VMs

[huddie]

We have a number of VMs here running in ESX4 (under vSphere). These particular VMs have been created using P2V from physical machines (not by me). One thing that all our P2Ved VMs have in common here is that when Windows Update Automatic Updates schedules an automatic installation which requires an automatic reboot, this reboot does not occur. Instead the VM goes part of the way through the reboot by beginning to shut down, then 'hangs'. It stays hung on the shutdown screen, not responding to the likes of terminal services / remote desktop, until I connect to it using the vSphere console. As soon as I connect and before I even do anything more, it then continues to shutdown and restart, but until I connect it will not do this. VMs have been known to stay hung like this for days.

I want to know how I can fix this problem so that when Windows Update Automatic Updates performs a scheduled install the consequent restart occurs successfully without me having to connect in.

Unfortunately, there are no specifics here. This issue occurs in our current versions - vSphere and ESX 4.0.0 - but it also occurred in previous versions that we had. I did not do the P2V conversion for these hosts. No errors are logged to the Windows Event Logs. Affected VMs are currently running WinXP SP3 but previous patch levels and service pack levels had the same problem. I have upgraded VMware tools to no avail.

Can anyone please help ? I'd be very grateful.

[proden20]

Never heard of this one. Is VMWare Tools installed and up to date? Do your Windows system logs show any errors? Are there any bad events in your VIclient on the VM?

Dennis Procopio - VCP

Please remember to award points for helpful or correct responses.

[huddie]

Hi Dennis.

Thanks for your reply. No, like I said, nothing in Windows event logs and VMware Tools has gone through different versions and is currently up-to-date.

What is the VLClient and how do I access it ? Sorry, but I've never heard of this.

--

Paul

[huddie]

Sorry Dennis, I've just realised you said VIClient not VLClient! 😕

I've checked and the vSphere client events and there's nothing there.

[huddie]

Dennis, I've attached the VMware.log file from the VM's datastore folder. There's some initialisation and service activity at the time the VM hung (20:04 11/11/2009) but nothing obvious to me. Thanks.

[proden20]

I've never seen anything like this but I'll throw more stuff out there..

- Can you post the .vmx file for one of the vm's

- Is there anything in the guest's device manager that the OS is expecting but no longer exists?

- there is a line in the log you posted about 3D being disabled but I'm not sure if that is normal. Is 3D enabled on your VM's video adapter? Maybe the fact that it reboots when you connect has something to do with a video driver.

Dennis Procopio - VCP

Please remember to award points for helpful or correct responses.

[huddie]

Thanks Dennis.

• The .VMX file is attached

• There shouldn't be anything expected but not configured in Device Manager, as nothing shows up as disabled or not detected.

• VM / Edit Settings / Hardware / Video card / Enable 3D support is unchecked and greyed out.

[proden20]

Running a VM optimally under vSphere requires both up-to-date VMWare Tools as well as Hardware Version 7. Your VM is on HW version 4. With the VM powered off, in the VM inventory, right click and upgrade the virtual hardware.

See if this helps, but lets start there.. There are methods to perform both Tools Upgrades and Hardware Upgrades in batch operations if you have that requirement as well.

Dennis Procopio - VCP

Please remember to award points for helpful or correct responses.

[huddie]

Dennis,

That's a great idea. I'd never thought of it. I've just upgraded the virtual hardware now.

The trouble is I can't reproduce this issue on demand. All I can do now is wait for the next MS Patch Tuesday and see if this causes it to hang again. I've tried initiating a reboot manually even before the hardware upgrade but, strangely, this does not trigger the problem. When I find out I'll let you know.

Thanks again.

[TobiasKracht]

after it you should also update VMTools.

StarWind Software R&D

[huddie]

Thanks Tobias, but I had already updated VMware Tools to the current version on our vSphere today, before the virtual hardware upgrade.

[huddie]

Thanks Tobias, but I had already updated VMware Tools to the current version on our vSphere today, before the virtual hardware upgrade.

[proden20]

If you are using a wsus server you can use "wuauclt /detectnow" at a command prompt. Select another mandatory patch in your WSUS server. If you are pulling direct from microsoft, try uninstalling a patch from add/remove programs or adding an old version of something like .net framework. Then run the command above.

That might help you reproduce the issue. I thought you had several servers that had this problem?

Dennis Procopio - VCP

Please remember to award points for helpful or correct responses.

[ScottBAnderson]

I also have the same problem described below. My virtual machines were converted using VMWare Converter 3.x, run both Server 2003 and Server 2008 OS and I have tried the following with no luck:

Convert virtual ide disks to SCSI. (LSI) using VMWare Converter 4.0.1 to import/export the entire system, also changed the disks from pre-allocated to not preallocated.

Uninstall VMWare tools, reboot, reinstall Tools.

Change Amount of RAM, change MPS version in virtual BIOS from 1.4 to 1.1, changed number of CPU's from 2 back to 1 and back.

Unlike the first poster, I can replicate the problem with ease. Initiating 'Restart' from within the guest OS always causes the hang. It happens right after power down when the console screen resizes.

I think this started with Windows Updates 3 or 4 months ago, so either July 14th or Aug 11th patch Tuesday. It worked fine when after I upgraded from server 2.0.0 to 2.0.1, then this problem came up and I updated to 2.0.2 hoping to resolve it, but like the other poster, this morning I had 5 VM's all stuck in the virtual BIOS screen (which is frozen hard, have to power off to boot it) and vmware-vmx.exe attempting to use 100% cpu each. Let me know what options you want me to try and what logs to obtain to help debug. I can tell you this much it hangs here in the standard log:

Nov 12 11:51:44.375: vcpu-0| PCIBridge7: ISA/VGA decoding enabled (ctrl 0004)

>>> It hangs right here on restart, boots just fine from power on, like the rest of this snippet shows: <<<

Nov 12 11:51:44.391: vcpu-1| CPU reset: soft

Nov 12 11:51:44.766: vcpu-0| CDROM-WIN: 'D:' Length 24 MaxTransferLength 131072 MaxPhysPages 32 ...

Now this is what I get when initiating the restart from within Windows.(and then I hit power off after it froze)

Nov 12 11:56:50.235: vcpu-0| PCIBridge6: ISA/VGA decoding enabled (ctrl 0004)

Nov 12 11:56:50.235: vcpu-0| PCIBridge7: ISA/VGA decoding enabled (ctrl 0004)

Nov 12 11:56:50.438: mks| VNCENCODE 2 encoding mode change: (720x400x24depth,32bpp,2880bytes/line)

Nov 12 11:56:50.469: mks| VNCENCODE 2 encoding mode change: (640x480x24depth,32bpp,2560bytes/line)

Nov 12 11:56:55.969: mks| SOCKET 2 (676) recv error 10054: An existing connection was forcibly closed by the remote host

Nov 12 11:56:55.969: mks| SOCKET 2 (676) destroying VNC backend on socket error: 1

Nov 12 11:56:58.250: mks| SVGA: display status changed, using optimizations for local consoles.

Nov 12 11:57:07.313: vmx| Stopping VCPU threads...

Nov 12 11:57:07.454: mks| Async MKS thread is exiting

Nov 12 11:57:07.563: vmx| MKS local poweroff ..

I've attached the entire log. The host is a dual xeon 5110 on a SuperMicro X7DBR-i+ (Intel 5000P chipset) running Server 2003 x64 with 16GB ram, a 3ware 9650SE-4LP adapter. I've also attached a vmx file

[tom11011]

I am having the same problem with ESX 4.0 and windows 2003. Did anyone find a solution to this problem?

[DSTAVERT]

You really should post your own question. Outline your problem with as much detail as you can. What steps have you tried etc. You can refer to other posts.