Data recovery - can I clone an entire ESXi Datastore?

[Desertsweeper]

So I restarted a host ESXi 5.1 server (after backing up the VM with Windows Backup) and it failed to boot off the USB drive with an error about insufficient buffers. No backup of the boot USB memory stick. Installed 5.5 to a new USB stick, created a new VM using existing disk and it booted fine. All seemed well until I started getting complaints about missing data. There was 10 days of data just gone. By the time I had discovered this the Scheduled Backup had overwritten my good backup. My last Veeam backup was on the day that missing data occurred. Discovered snapshot fragments indicating an incomplete backup.

So to summarize, I created a new VM with a previous VM's VMDK that had not been consolidated.

Now I want to try recovery tools but it is taking ages. I have a standalone Core-i7 PC with 2 x 2TB SSD drives in Raid-0. I want to copy the entire Datastore from the host to this box so I can run recovery scans at a decent speed.

Is this possible?

Details: HP Proliant DL380p G8 with 4x1TB SATA Drives in Raid5

ESXi 5.1 was installed on USB drive

ESXi 5.5 installed after failing to boot

[continuum]

> Is this possible?
Yes. I do this quite often.
But it only makes sense if you clone the raw device itself with dd or ddrescue.
Would you mind and give a few more details ?

[Desertsweeper]

Hi continuum, thank you for responding.

Cloning the raw device would mean booting the server off a cloning application disk right? Then copying the raw disks to an attached drive or across a network. Something like Acronis? But what recognises the VMware partition system?

What sort of details would you like? It is as though the VM rolled itself back to the 2nd of may, 10 days prior to the issue.

[continuum]

> It is as though the VM rolled itself back to the 2nd of may, 10 days prior to the issue.
That is usually the result of using a misconfigured vmx-file.
You probably assigned the wrong snapshot. Stop using the VM with outdated state as soon as possible.

[continuum]

I dont see a need to clone anything at this moment - it may be as simple as assigning the right snapshot.
Please provide a filelist of what you have inside the VMs directory.

[Desertsweeper]

Thank you for your continued assistance continuum. Here is what is in the VM Folder:

There are two volumes, ServerDXB2.vmdk and ServerDXB2_1.vmdk

[Desertsweeper]

I only just saw your advice on the misconfigured vmx file and to stop using the outdated VM. Unfortunately the users have been churning away for the last few days...

[continuum]

I think this should be easy to fix - unless the damage you have created by the misconfigured vmx and vmdks is already too severe.
If possible call me via skype and let me do the repairs remotely - should take just a few minutes.
See my signature

[Desertsweeper]

Hi continuum, is it too late to call you?

[continuum]

Nope
fastest way to contact me is via skype - feel free to ask questions or request a phonecall any time
Ulli

[Eiber Azuaje]

Maybe a bit too late to ask, but did you managed to rescue your datastore? I'm in a similas situationas where my disk has two VM's, the first one works flawlessly but the second one says it's not valid. Browsing arround I noticed a .lck file and I can't copy any out of that datastore to try making a new vm attaching .vmdk flat.

Bottomline, any advice will help. Thanks.

-------------------------------------------
Mensaje original:
Enviado: May 17, 2016 02:03 PM
De: Desertsweeper
Sujeto: Data recovery - can I clone an entire ESXi Datastore?

Hi continuum, thank you for responding.

Cloning the raw device would mean booting the server off a cloning application disk right? Then copying the raw disks to an attached drive or across a network. Something like Acronis? But what recognises the VMware partition system?

What sort of details would you like? It is as though the VM rolled itself back to the 2nd of may, 10 days prior to the issue.

[ninjabrum]

trying to revive a 9 year old thread won't get you far.  Start a new topic with more details that what you provided (versions, steps taken to troubleshoot).  If it's locked, then the host thinks it's running.  Try restarting esx management agents or rebooting the host.  There are multiple hits when I search for "locked VM".