VMware Horizon Community
Petrucci914
Contributor
Contributor

View 4.6 - Link Clones Pool stuck at 'Customizing'

Hello,

We have View 4.6.  I have created several Link Clones Pools running Windows XP.  There is one pool where I keep running into issues.  I initially created it for 16 desktops.  It would deploy all of them, I'd see all the accounts created in AD, but many of the desktops would stay at 'customizing.'  When I would open up their console it would be at the beginning of the Windows XP Setup Screen.  But other desktops in the pool created just fine and worked perfectly.  I have other pools using the same Sysprep file and those are created fine.  I've tried creating the pool with lesser desktops and I still have the issue.  I've also tried re-creating the Master Image Snapshot on which it is based but nothing has changed.  After they eventually error out I'll delete them so that it re-creates them and sometimes it creates more successfully but there are always a few that fail.  This pool was created successfully before so I'm not sure why it is causing issues now.....

Any Ideas?

0 Kudos
35 Replies
idle-jam
Immortal
Immortal

why not use quickprep instead? i find it faster and more reliable ..

0 Kudos
Petrucci914
Contributor
Contributor

I use the VMWare Customization Wizard to create the Sysyprep process, so it's not a Sysprep file I had created before.  This method works with other pools.  All are Windows XP SP3.

0 Kudos
tacticsbaby
Expert
Expert

I had a similar problem in my View environment. I would provision a pool of linked clones and some of them would never finish the customization process. In my case this problem was caused by a number of issues:

1) Insufficient IP addresses for the pool. Unfortunately customization errors can be caused many things!

2) Just a weird glitch or perhaps bug in View (I am currently using version 4.0.1) - In this case I just reboot the VM and customization works.

3) Authentication failure for quickprep/sysprep account. I have had this issue when the AD account did not have access to the OU structure in which the VMs resided.

These are just a few of the reasons we got this message in our environment. If it turns out that this problem can be solved by a reboot then you may wish to add a post syncronization script to your VMs that reboots them once the provisioning process is finished, no matter what. Hope this helps.

0 Kudos
Petrucci914
Contributor
Contributor

Thanks for the response.

We have enough IP Addresses in that scope.  If I reboot the VM it still gets stuck at the same point.  The OU has the correct permissions as other pools have no problem creating accounts there.  It also shows all accounts there from the problem pool, the VMs just never advance to completion.

0 Kudos
tacticsbaby
Expert
Expert

In that case have you checked your DNS. That was also a problem with us. Truly an interesting problem.

0 Kudos
Petrucci914
Contributor
Contributor

Good call.  I see that the Computer Accounts are created in the OU but there are no A Records associated with them in DNS.  I have the VIEW account set up for permissions on the OU.  Anything else I may have to check to make sure it has permissions to fully create the accounts and ensure an A Record is created?

0 Kudos
tacticsbaby
Expert
Expert

One thing that worked for me was to release the IP address of the master image before shutting it down and snapshoting it. Hope this helps.

0 Kudos
Petrucci914
Contributor
Contributor

Ya, I've tried that as well.

This is getting annoying.  Everything was working fine and nothing changed as far as I know.  Hulk angry.  Hulk smash!

0 Kudos
tacticsbaby
Expert
Expert

Another idea that helped me was to just delete from View and if needed virtual center the VMs that never customized. Completely remove them from AD, VC and View. First try to remove them from the View manager. I doubt this will work because they are not finishing the customization process. Once you delete the VMs then just reprovision them. That should work. If not, then Hulk SmashSmiley Happy

0 Kudos
Petrucci914
Contributor
Contributor

Ya, I've already done that as well.

....time to get new purple pants.

0 Kudos
tacticsbaby
Expert
Expert

Just for sits and giggles have you tried to manually create the A records for these VMs? I know this is not a real solution, but it may help shed light on this problem. This also worked for me. Just out of curiosity, what versions of VC and ESX are you using? What hardware version are you using for your VMs?

0 Kudos
Petrucci914
Contributor
Contributor

I did manually create a few but no luck.  If I reset the VM then the size of the VM changes minutely but it still stays at that step.  I then started looking through the datastore that it uses and saw duplicates of these VMs (denoted by _{n}).  I tried to delete the pool but no luck, so I deleted the VMs from the Datastore.  Now when I try to clear out the Pool or VMs from the View console I get an error because the Datastores for them are no longer there.

Wow, what a mess this has become.

ESX, vCenter, and View are all the latest versions.

0 Kudos
tacticsbaby
Expert
Expert

Are you able to login to any of these failed VMs? If so then perhaps you may be able to get a better idea of what is going on. Sorry to hear this is such a pain. When we went through this, like I said it was due to a long list of reasons. In my case the root cause was a problem with the base image. We just created a new master image and made a few tweaks to our environment like adding more IP addresses (I know this is not your problem), cutting the lease times down to 1 hour and redeploying the pool from the new base image. Even now when refreshing or recomposing 20 or more VMs some of them will not complete the customization process. In our case a simple restart of the VM will fix it.

0 Kudos
ayang
Contributor
Contributor

I am also having the same issue.  LinkClone Automated Floating Pool.  Only half VDIs are running, the other half won't finished the customization.  Even we manually reboot the troubled half, they do came up & joined to the domain, but the VDI pool don't recognize them

I put in a SR.. but no response yet..

Anyone has any suggestion?  Looks like VDI & View is like a step child at VMWare....

Help!

0 Kudos
Petrucci914
Contributor
Contributor

Ya, I'm still stuck as well.  I'm giving the server a reboot before I contact VMWare support.

I browsed the Datastore and saw that there were numerous duplicate bad VMs.  I removed all of those and then I couldn't remove the Pools or VMs from the View Console so I had to use ADSI and remove the data that way.

I'm thinking it is some bug in 4.6.  This VDI stuff is becoming laughable....Should have gone with Citrix.

0 Kudos
tacticsbaby
Expert
Expert

Did you check to make sure you have enough ports on your vswitch?

0 Kudos
riki78
Contributor
Contributor

Hello

We have also problems with vm's in cutomizing state...

What kind of antivirus you use?

Best regards

Simon

0 Kudos
tacticsbaby
Expert
Expert

We use McAfee VSE 8.7. We only use the virus scanner and do not use firewall or ids. If you are using any additional features of your virus scan this may be an issue, but the fact that some of your VMs are working properly suggests that this is likely not the cause.

0 Kudos
kgsivan
VMware Employee
VMware Employee

If the agent is XP and Domain controller is on 2008 then it will be due toa known issue.

Before taking the snapshot please install a microsoft patch from here

0 Kudos