MSCS adding virtual quorum resource

[MarshMallowEgg]

Hi,

I'm running MSCS 3.5 and Windows Server 2003. I'm attempting to create a two node cluster on a single physical host. I've followed the instructions provided in the various pdfs but for some reason I cannot bring the virtual quorum online in cluster administrator. So far I have two nodes, each one with two scsi controllers (both set to virtual), a local disk, a shared quorum drive (zeroed, then an extended partition + logical partition created, also independent+persistant, a shared data drive (same settings as quorum).

I add the resource, which goes fine. I then try to bring it online and it fails. Does any one have any ideas about what I need to do? Do I need to disable the disk locking?

Thanks

[Texiwill]

Hello,

Make sure the quorum drive is of type Thick or Zeroed Thick.

Best regards,

Edward L. Haletky

VMware Communities User Moderator

====

Author of the book 'VMWare ESX Server in the Enterprise: Planning and Securing Virtualization Servers', Copyright 2008 Pearson Education. As well as the Virtualization Wiki at http://www.astroarch.com/wiki/index.php/Virtualization

[MarshMallowEgg]

I tried creating a new disk using zero thick and adding it, unfortunately it still fails when I try to bring it online.

[kjb007]

Look in the vmware.log file right after you try to bring the disk online. There may be some other underlying issue. That file should be located with your vm files.

-KjB

[MarshMallowEgg]

I looked in the log after my attempt, doesn't look like anything is wrong. I'm attaching it in case I've missed something:

May 08 10:40:54.380: vcpu-0| VMXVmdbGuest_GetGuestUptime failed. ret = Type mismatch

May 08 10:40:54.380: vcpu-0| DISKUTIL: scsi1:0 : toolsVersion = 7299

May 08 10:40:54.380: vcpu-0| DISKUTIL: scsi0:1 : toolsVersion = 7299

May 08 10:40:54.380: vcpu-0| DISKUTIL: scsi0:0 : toolsVersion = 7299

May 08 10:40:54.380: vcpu-0| TOOLS installed version 7299, available version 7299

May 08 10:40:54.380: vcpu-0| TOOLS don't need to be upgraded.

May 08 10:40:54.448: vcpu-0| Guest: toolbox: Version: build-64607

May 08 10:40:54.448: vcpu-0| TOOLS unified loop capability requested by 'toolbox'; now sending options via TCLO

May 08 10:41:29.508: vmx| TOOLS received request in VMX to set option 'synctime' -> '0'

May 08 10:41:29.522: vmx| VMXVmdb_LoadRawConfig: Loading raw config

May 08 10:41:40.956: vcpu-0| HBACommon: First write on scsi1:0.fileName='/vmfs/volumes/725d7b94-b09d0694/MSCS-Test1/quorum.vmdk'

May 08 10:41:40.960: vcpu-0| DISKLIB-CHAIN : UpdateContentID: old = 0x252be49d, new = 0xec574a49

May 08 10:41:59.346: vmx| TOOLS received request in VMX to set option 'synctime' -> '0'

May 08 10:41:59.351: vmx| VMXVmdb_LoadRawConfig: Loading raw config

May 08 10:42:29.585: vmx| TOOLS received request in VMX to set option 'synctime' -> '0'

May 08 10:42:29.591: vmx| VMXVmdb_LoadRawConfig: Loading raw config

May 08 10:47:59.946: vmx| TOOLS received request in VMX to set option 'synctime' -> '0'

May 08 10:47:59.965: vmx| VMXVmdb_LoadRawConfig: Loading raw config

May 08 10:53:30.335: vmx| TOOLS received request in VMX to set option 'synctime' -> '0'

May 08 10:53:30.356: vmx| VMXVmdb_LoadRawConfig: Loading raw config

May 08 10:58:47.895: vcpu-0| MKS Backdoor get pointer: first time, notify tools are running

May 08 10:58:48.049: vcpu-0| TOOLS unified loop capability requested by 'toolbox-dnd'; now sending options via TCLO

May 08 10:58:48.105: vcpu-0| GuestRpc: Channel 2, registration number 1, guest application toolbox-dnd.

May 08 10:58:48.105: vcpu-0| DISKUTIL: scsi1:0 : toolsVersion = 7299

May 08 10:58:48.105: vcpu-0| DISKUTIL: scsi0:1 : toolsVersion = 7299

May 08 10:58:48.105: vcpu-0| DISKUTIL: scsi0:0 : toolsVersion = 7299

At this point it might be an issue with Server 2003 or cluster admin itself.

The only error in the windows event logs is an "unspecified error 0x80004005" x 4. From things I've looked at on the internet there have been similar issues but those were solved by people reinstalling/reconfiguring actual physical hardware solutions.

I also read something about the volume labels but I'm not sure what the person was getting at. If any one can answer this though, when I mount the virtual disk, should I assign drive letter? should I format it in NTFS? should I apply a volume label?

[kjb007]

While you're trying to bring online, is your 2nd node offline? Just make sure you get everything online on one node before adding the 2nd. When I've had cluster issues, it's usually because the 2nd node is trying to do something it should not.

-KjB

[MarshMallowEgg]

The 2nd node is offline. But now that I think about it, I have tried mounting it when the 2nd node was online. Is it possible that a lock was created on the disk the has not been cleared?Is there a way to clear locks?

Also, I found another symptom, the disk is not accessible normally by both. I read that if I make a change on the disk it should be visible on from both nodes. Both nodes can "see" the disk, but if I create a folder for instance, it's only visible on the node that made the folder.

BTW, right now in both node's vmx files I have set disk.locking ="false". It was something I saw from the forums before that I wanted to try. Didn't seem to change anything...

[kjb007]

Without clustering, only one node will see changes from the time a drive letter was given to it and it was mounted. For instance, if you mount on node 1, give drive letter, and make changes, then on node 2, mount and give drive letter. The changes from node 1 will be seen. But, now that you have mounted, node 2 will not see anymore changes from node 1. This is because both nodes are thinking they own the drive, and are not aware of anything else having access to that drive. The easiest way to clear a lock, is to reset the node you want to clear it from. If it is already down. Then, I would reboot node 1, and retry changes on node 2. Also, make sure the time is in sync between both nodes.

-KjB

[kjb007]

I just found this article :

Did you add the shared disks before adding the NIC devices? Try using dumpcfg to see if you need to reset the Disk IDs.

-KjB

[Priergaard]

Hi!

I have the same problem...any solution yet....???

Regards Mogens