VMware major issue on Christmas - Hating vm !!

[ranjitcool]

Hey Guys,

Not a great start but hoping to get some expert advice during this holiday! Below is the sequence of events.

1. vmware esx 3.5 server standalone connected on fiber to san

2. did a cold copy move of a linux vm from local disk to SAN datastore

3. Started the vm and started well with networking issues

4. I panicked and moved it back from SAN DATASTORE to local datastore

5. networking issue remained, looked online, found solution fixed it. Started to ping. Vm is now on local datastore

6. got a error when i tried to restart - Failed to open '/vmfs/volumes/4a365b5d-eceda119-439b-000cfc0086f3/examplevm/examplevm-000002.vmdk' with flags 0xa (The parent virtual disk has been modified since the child was created).

7. Looked online - its a cid issue so tried to fix it, file examplevm-00001.vmdk improperly linked to examplevm.vmdk - fixed the parentcid - removed vm from inventory and added it back as the kb article said - no joy

8. VM boots but none of my data is there, it seems to revert to the first snapshot.

9. No disk writes or changes were made, just multiple boots to trying to check if i got my files back. Even snapshot rollback fails, it creates a examplevm-0000X.vmdk file - complains it cant open the disk. X here is an increasing number.

10. VMWARE said a bug exsits - http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=100522...

11. Now making copy of datastore to local desktop to bump vm to esx 4.1

My question - can i recover my vm/data ? Will santa grant my only christmas wish?

Thanks

RJ

[ranjitcool]

bump!

[Linjo]

Hi.

So if I understand correctly there was one or more snapshots on the vm from the beginning?

Could you post the vmx-file and a file listing of all the files in the vm directory?

// Linjo

[ranjitcool]

Thanks for writing back, i looked at vmx, it says its deployment - windows but its a linux box set to redhat 64 bit.

I am confused is that the reason ??

Yes there were about 4 snapshots with the last one pon august 09th 2010.

These are the files in the vm directory

HCTC_Websrv_hctc.net.vmxf

hctc_webserv_hctc.net.vmsd

hctc_websrv_hctc.net.vmx

hctc_websrv_hctc.net.nvram

hctc_websrv_hctc.net-snapshot9.vmsn

vmware-28.log

vmware-27.log

vmware-29.log

vmware-31.log

vmware-32.log

vmware-30.log

vmware.log

hctc_websrv_hctc.net-000006.vmdk

hctc_websrv_hctc.net-000007.vmdk

hctc_websrv_hctc.net-000008.vmdk

hctc_websrv_hctc.net-000010.vmdk

hctc_websrv_hctc.net-000009.vmdk

hctc_websrv_hctc.net-Snapshot2.vmsn

hctc_websrv_hjctc.net-snapshot4.vmsn

hctc_websrv_hctc.net-snapshot5.vmsn

hctc_websrv_hjctc.net-shnapshot6.vmsn

hctc_websrv_hctc.net-snapshot1.vmsn

hctc_websrv_hctc.net-000001.vmdk

hctc_websrv_hctc.net-000003.vmdk

hctc_websrv_hctc.net-000004.vmdk

hctC_websrv_hctc.net-000005.vmdk

hctc_websrv_hctc.net-000002.vmdk

hctc_websrv_hctc.net.vmdk

the above is the file list.

thanks

RJ

[ranjitcool]

Also in vmx file it says

scsi0:0.fileName = "HCTC_Websrv_hctc.net-000010.vmdk"

shouldnt that be hctc_websrv_hctc.net.vmdk? i compared it to another vmx file. Also 000010 is a 1mb file

This is weird i think. I could be wrong.

RJ

[Linjo]

Ok, so you probably have some messed up cid-chains, the original problem was most likely that you moved the vm without the snapshot files.

I would recommend to read this carefully and follow the instructions listed: http://sanbarrow.com/sickbay.html

// Linjo

[ranjitcool]

Yes you are right i have messed up cid chains.

What i am thinking of doing is to visit the site and load vmware 4.1 as this 3.5 is an old machine with no updates.

Once 4.1 is in i want to connect it to the SAN where i backed up a copy of what i have and add it to the inventory and try.

What do you think of this?

I will look at the article, thanks. I tried the kb article from vmware - no joy

RJ

[ranjitcool]

Quick update, just like that article and the kb article earlier - i spotted the differennt parent cid in the main .vmdk file and changed it \

tried to start the vm and no joy. checked vmdk file again and it has another id, different than the one i assigned.

Does this explain? - http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=100522...

[a_p_]

Yes there were about 4 snapshots ...

You don't have only 4 snapshots, you actually have 10 of them.

4. I panicked and moved it back from SAN DATASTORE to local datastore

Next time you have an issue, stop at this point and open a thread. Don't try to fix things by yourself, if you are not 100% sure how things work together!

To be able to help you, let's start from the beginning. Please answer to following questions and attach the files (even  the ones already attached) needed to be able to find out what can bee done:

• a complete list of files in the datastore showing names, sizes, time/date. The above list does not contain the flat and delta files!?
  
• the current vmx file
  
• the current vmsd file
  
• all log files
  
• all header vmdk files (the ones with only a few bytes in size)
  

• How did you copy the VM from the local to SAN datastore? I guess you used the datastore browser.
  
• Which vmdk files did you edit so far (repairing the CID chain)?
  
• How much free disk space do you have on the datatstore?
  
• Do you have other datastores with enough free disk space to clone the current virtual disk?
  
• Did you edit anything else?
  
• Are you familiar with the ESX hosts's console?
  

André