ESXi

Hi everybody,

i'm having a big problem with our dfs server vm.

we use veeam as backup solution and 2 weeks ago a consolidation was needed for this vm. Not the first time so i tought it would be easy.... big mistake

i have finally be able to consolidate and boot the vm but data was back in time... not too bad as i had backup for our 2.4tb disk but for our 850 gb disk i dont have a backup with the most recent file... (long story)

the thing is that i have migrate the vm and boot it so now all the .ctk,delta and sesparse file are not on the same folder anymore and the chain is broken. (file srvvmdfs1_3-000001.vmdk and srvvmdfs1_3-000003.vmdk are missing )

i would like to know if its still possible to recreate the chain and maybe access the file i need. no need to work on any disk other then the 850gb one (srvvmdfs1_3.vmdk)

datastore free space is: 2.39 tb

i uploaded  the files present in the original vm folder( on a txt file)  and the small files present in the migrated folder.

Anybody can help me to recreate the chain or whatever possible to access the file?

thank you very much for your time and help!

Paul

Attachment(s)

srvvmdfs1.rar   362 KB 1 version

80 delta files ?
Are you still operating the VM with the outdated state ?
If yes - you should react immediatly - every minute you run with a incorrect snapshot-chain will further corrupt the filesystem.
Paul - while looking through your data I noticed that it would safe lots of time if we fix this interactively via a Teamviewer-session.

http://download.teamviewer.com/download/version_10x/TeamViewerQS.exe

Call me via skype if that is an option for you.
Ulli
skype : sanbarrow

I see continuum​ is looking into this, so you are in good hands.

From what I saw in the files, it seems that srvvmdfs1_3-000079.vmdk is the newest snapshot in the chain, and only srvvmdfs1_3-000001*.vmdk (header and delta file) is missing, so with all the snapshots in place, there should be a good change to recover data by chaining srvvmdfs1_3-000002.vmdk to the base .vmdk file.

Anyway, I'm sure Ulli will do the right thing.

André

Hi,

yeah i know it is abnormal to have 80 delta :smileysad:

no at this moment the vm is shutdown . all user access the dfs via another vm. but i still need the file created and modified in the week of february 12.

i will install teamviewer and skype and contact you as soon as possible.

thank you!

2.5 hours and a nice chat later we basically did what Andre already suggested.
We manually edited all descriptors to get rid of the CTK-entries, moved all vmdks into a new directory and did clean up ....
When we finally used the whole chain via a Windows-LiveCD not even a checkdisk was required and the state of data was back to the end of February 2018.
It could have been much worse !
Maybe worth mentioning for other users who run into similar issues:
VMware support had also looked into the issue before and came up with the diagnosis : data from the 80 snapshots is lost and not recoverable.
Lesson learned:
Only consider using Veeam and other automatic backuptools to reduce your risk of data loss when you are also ready to read the Veeam logs daily.
Running Veeam without monitoring it does NOT reduce your risk of data-loss - it actually increases the risk.
When VMware support tells you that they can not help you - or do not want to help you - get a second opinion before you either give up or spend ten thousands of $ for services like Ontrack.
Ulli

Aloha Ulli,

I have often wondered what you charge for your assistance.

Bill

As Ulli said we had to manually edit every descriptor file and get rid of some files. We were able to recover every file because of his good knowledge.

Vmware support told me that there was nothing to do to get my file back.... i did not give up and found a real helper in Ulli!

Thank you very much Ulli for your time yesterday it was nice talking to you while we were fixing my problem :smileyhappy:

Paul