VMware Horizon Community
Ybbom
Contributor
Contributor
Jump to solution

VDI client stops during recompose process

Hello

We have quite a few VDI pools and therefore a lot of VMs (Windows 7 Enterprise x64) running on it (Horizon View 6.1 on vSphere 5.5 Update 2d). We recognised that some VMs not properly finishes the recompose process and it happens completely random (most of the time there were only 1 up to 5 VMs affected from these effect). There are no related errors shown in the log files and the event viewer on the VDI machines also does not contain any information. When I compare the log files and the task history, it seems that everything is working properly. The result is a shut down VM with status "Agent unreachable" and some warning entries in the View Administrator event log about users who tries to login on the VM but fails.

 

One error that I found was about an unexpected exception and the IP address cache on the VM itself:

2016-06-02T16:03:33.171+02:00 ERROR (0620-0894) <TopicMessageManager-agent> [AbstractJmsConnectionUser] Hit an unexpected exception, requesting clean reconnect

2016-06-02T16:03:33.171+02:00 DEBUG (0620-0848) <2120> [wsnm_jms] AddressCache::RefreshCache: No addresses are currently set

2016-06-02T16:03:33.171+02:00 DEBUG (0420-04DC) <1244> [wsnm_desktop] AddressCache::RefreshCache: No addresses are currently set

 

The VM get cloned from the master image and after this process some reconfiguration tasks take place. The only strange thing during the recompose is, that the VM tries to join the active directory but fails about 1-3 times before it successfully completes this task:

 

2016-06-02 16:01:50,187 [2724] INFO  SvmGa  - [svmGa.cpp, 711] Domain join failed with 18

2016-06-02 16:01:50,203 [1692] DEBUG Wow64FsRedirectionOff -  [Wow64FsRedirectionOff.cpp, 135] Successfully reverted WOW FS redirection.

2016-06-02 16:01:50,203 [1692] FATAL VolumesReady -  [VolumesReady.cpp, 135] Joining Domain failed for 2 times.

2016-06-02 16:01:50,203 [1692] DEBUG TimerWrapperWin32 -  [TimerWrapperWin32.cpp, 47] Starting timer: DomainJoinRetryTimer-Handle with timeout: 60000 milliseconds

2016-06-02 16:02:50,203 [1676] INFO SvmStateMachine  -  [SvmStateMachine.cpp, 74] Queing an event

2016-06-02 16:02:50,203 [1676] DEBUG WaitableManagerWin32 -  [WaitableManagerWin32.cpp, 179] Dispatching callback for event DomainJoinRetryTimer-Handle (000001DC)

2016-06-02 16:02:50,203 [1692] INFO VolumesReady  -  [VolumesReady.cpp, 115] Launching the Join Domain script: C:\Program Files (x86)\Common Files\VMware\View Composer Guest Agent\vmware-svi-ga.exe /JoinDomain

2016-06-02 16:02:50,203 [1692] DEBUG Wow64FsRedirectionOff -  [Wow64FsRedirectionOff.cpp, 109] Successfully disabled WOW FS redirection.

2016-06-02 16:02:50,250 [1620] INFO CSvmGaService  -  [svmGaService.cpp, 87] Initialized the Service

2016-06-02 16:02:50,250 [1620] INFO CustomizationPersistentInfoReg - [CustomizationPersistentInfoReg.cpp, 98] Opened guest registry key for accessing customization info.

2016-06-02 16:02:50,250 [1620] DEBUG SvmDisk  -  [SvmDisk.cpp, 55] Found signature file //?/Volume{e4f154e4-28c9-11e6-a287-806e6f6e6963}/simvol.dat at volume \\?\Volume{e4f154e4-28c9-11e6-a287-806e6f6e6963}\ with mount point * and usage 001

2016-06-02 16:02:50,250 [1620] INFO CSvmGaService  -  [svmGaService.cpp, 214] Found policy file on the volume: \\?\Volume{e4f154e4-28c9-11e6-a287-806e6f6e6963}\

2016-06-02 16:02:50,250 [1620] INFO CSvmGaService  -  [svmGaService.cpp, 305] Successfully parsed the policy and disk signatures

2016-06-02 16:02:50,250 [1620] DEBUG CSvmGaService -  [svmGaService.cpp, 131] Joining domain

2016-06-02 16:02:50,250 [1620] INFO  Guest  - [Guest.cpp, 296] Attempting to join VIRTUALMACHINE to the domain DOMAIN.TLD

2016-06-02 16:02:51,718 [1620] INFO  Guest  - [Guest.cpp, 313] Domain Join successful

 

When I check the VM in the vCenter Server console the VM is shut down and has no IP address assigned to it but a vmxnet3 adapter is still configured. If I manually power on the VM, an IP address gets assigned to it and the reconfigure process finishes successfully. After a reboot the VM becomes available to the VDI pool and the error message in the View Administrator event database disappears. But why is there a problem for the VM to get an IP address? We checked the DHCP server but there is also no problem and we also have more than enough IP address for the VDI pool. I also tried the https://kb.vmware.com/kb/2007319 but the problem still exists and it happens randomly on quite a few VMs.

Had someone the same issue?

Kind regards,

Ybbom

 

Reply
0 Kudos
1 Solution

Accepted Solutions
Ybbom
Contributor
Contributor
Jump to solution

We upgraded to Horizon 7.1 and the issue seems solved.

View solution in original post

Reply
0 Kudos
2 Replies
BigDigital
Contributor
Contributor
Jump to solution

Hi Ybbom,

I've had similar issues in the past which were resolved by uninstalling the View Agent and VMTools.  Reboot and re-install VMtools and View Agent.  I run ipconfig /release along with ipconfig /flushdns before the snapshot capture.   Hope this helps!

Reply
0 Kudos
Ybbom
Contributor
Contributor
Jump to solution

We upgraded to Horizon 7.1 and the issue seems solved.

Reply
0 Kudos