Cannot Power On VM - General Error

[swinster]

Hi All,

After up dating a VM, I shut down the VM and Host. On start up, I recieve the a couple of errors.

-----------------------------------------------------------

Event Details:
	Type: error Description: Error  message on My.VMl on My.Host in  ha-datacenter: We will respond on the basis of your support entitlement. Time: 02/09/2012 17:02:17

Error Stack:
	We will respond on the basis of your support entitlement. To collect data to submit to VMware technical support, run "vm-support". You can request support and include the contents of the log file. A log file is available in "/vmfs/volumes/4f4aae60-f73239c8-aea8-00e081ca7d70/My.VM/vmware.log". VMware ESX unrecoverable error: (vcpu-0) Cannot start VCPU thread 1 Cannot start thread: 0 (Resource temporarily unavailable)
Additional Event Details:
	Type ID: Error Host Build: 469512 Target Object Type: ComputeResource DataCenter Object Type: Datacenter Host Object Type: HostSystem VirtualMachine Object Type: VirtualMachine ChainId: 8

-----------------------------------------------------------

Cannot power On
My.VM on
My.Host. A general system error
occurred:
error
02/09/2012 17:02:22

-----------------------------------------------------------

This is a home based ESXi so I have no support to be able to submit the error report. Does anyone have any ideas?

I have tried creating another VM and pointing it to the VHDD of the the failing VM, but I still see a similar error -

-----------------------------------------------------------

Event Details:
	Type: error Description: Error message on NewVM on My.Host in ha-datacenter: We appreciate your feedback, -- the VMware ESX team. Time: 02/09/2012 17:37:51
Error Stack:
	We appreciate your feedback,   -- the VMware ESX team. We will respond on the basis of your support entitlement. To collect data to submit to VMware support, run "vm-support". If the problem is repeatable, set 'Use Debug  Monitor' to 'Yes' in the 'Misc' section of the Configure Virtual Machine  Web page. Then reproduce the incident and file it according to the  instructions. You can report this problem by selecting menu  item Help > VMware on the Web > Request Support, or by going to  "http://vmware.com/info?id=8&logFile=%2fvmfs%2fvolumes%2f4f4aae60%2df73239c8%2daea8%2d00e081ca7d70%2f...".  Provide the log file  (/vmfs/volumes/4f4aae60-f73239c8-aea8-00e081ca7d70/NewVM/vmware.log) and the core file(s)  (/vmfs/volumes/4f4aae60-f73239c8-aea8-00e081ca7d70/NewVM/vmmcores.gz, /vmfs/volumes/4f4aae60-f73239c8-aea8-00e081ca7d70/NewVM/vmx-zdump.000). *** VMware ESX internal monitor error *** vcpu-0:PCIPassthru: failed to adjust IOMMU mappings.
Additional Event Details:
	Type ID: Error Host Build: 469512 Target Object Type: ComputeResource DataCenter Object Type: Datacenter Host Object Type: HostSystem VirtualMachine Object Type: VirtualMachine ChainId: 26

-----------------------------------------------------------

Is there anyway I can retrieve the VM on the VHDD in the datastore?

Chris

[onoski]

The errors presented on the table you attached is very generic, is there a chance you can post the exact error or the log file pertaining to this issue?

[swinster]

Attached if the vmware.log from the initial error. I have just deleted the tempoary VM from the datastore but could always recreate if needed.

[onoski]

I have just gone through the log file and couldn't find any relevant information or errors. If you don't mind recreate the VM and make sure it has VMware tools installed and try again.

Best wishes

[a_p_]

From the log file it seems you have an USB device configured as passthrough device in the VM. Is this device still present after the host reboot? I'd suggest you try to disconnect the passtrough device from the VM and try to start the VM again.

André

[swinster]

I have looked over the log myself and ineed noted the something relating to a PCI passwthough device. I DO have an Adaptec RAID controller that is set in passthough mode. - this is to allow the VMs access to the main data volume. As this is all contianed on the one host, this is the only way I could achive this.

I can disable the passthough on this particular device and test to see of the VM boots, but the VM wont have full functionality.

---

OK, just remove the device from the VM, tried to restart and I still gen a "General Error". So I removed the PCI device completely as a passthrough device from the host, then tried to reboot from the vSphere client but also recieved a "General error". Hmm.

Maybe somehting has gone screwy with the ESXi install. I was going to perform an inplace upgrade to 5.0 Upgrade 1 - maybe I should have a go at doing this to refresh the install?

[a_p_]

I was going to perform an inplace upgrade to 5.0 Upgrade 1

Out of curiosity. Which version/build was installed before and did you also install the latest 5.0 patches?

André

[swinster]

The previous install was 5.0.4 somthing. No patches were installed - everyhting was working OK.

Other oddities - can't reboot or shut down the host directly. Had to physically power down the machine.

[swinster]

Looks like some of the setting in the ESXi host were corrupt with regard to the passthrough devices. I tried an upgrade, but that didn't work. I have now re-installed ESXi and the passthough Adaptex RAID controller is recognised again. Of course, I have not lost all my other configuration including VMs' I think there shoudl be a way to pull in the VM data from the datastore, but I'm going to have to look this up.

[swinster]

OK, someway back to a functioning system. Looks like the refresh install has worked. Added the VM back into the inventory, with passthough to the Adaptec RAID controller, and all appears to be functioning.

I'm sure there is a way to save some configured preferences such as resource pools, network config etc, but in this case I think it relatively simple to re-configure from scratch. However, if anyone has got any info on how to save these for another time, I would appreciate it.

[swinster]

This issue looks like ESXi had got kinckers in a twist about the hardware on the host - is there anyway to refresh this without a complete re-install of ESXi - just for future refernce?