VMware Communities
ChrisHasenpflug
Contributor
Contributor
‎11-05-2008 08:31 PM

Recreating vmdk file

So I was an idiot and accidently ended up deleting my main .vmdk file. This is a 20GB sparse disk with 2GB parts. All of my actual .vmdk files are there (s001 - s010) and I have some snapshots too, I just ended up deleting the main .vmdk file. From looking at some of the snapshot .vmdk files, it looks this is just a text file with information about the sparse disk parts. One would think I'd just be able to re-create this file, but I'm not quite sure what all needs to be there.

From looking at a different virtual machine, I'm thinking my parentCID would be "ffffffff" since it's the main image, not a snapshot, right?

What is the CID? Would I need to look at my first snapshot's parentCID? Just infering here, so if there's an easier way please let me know.

I see the snapshot's have a "parentFileNameHint"...is that necessary for the main disk files?

Also noticed my other machines have alot of ddb... entries, whereas the snapshots only have ddb.encoding. Does it matter?

I'm just trying to get this machine to boot back up. It's actually about to be scrapped, but I need to pull a few files off it.

Thanks!!

0 Kudos
1 Reply
admin
Immortal
Immortal
‎11-10-2008 04:43 PM

Sorry for the delay in responding.

From looking at a different virtual machine, I'm thinking my parentCID would be "ffffffff" since it's the main image, not a snapshot, right?

Correct.

What is the CID? Would I need to look at my first snapshot's parentCID? Just infering here, so if there's an easier way please let me know.

Yep. I think it's short for Chain ID or something - basically, it's a sanity check that the parent virtual disk is the expected one.

I see the snapshot's have a "parentFileNameHint"...is that necessary for the main disk files?

I don't think it's necessary in general, actually - as the name implies, it's a hint.

Also noticed my other machines have alot of ddb... entries, whereas the snapshots only have ddb.encoding. Does it matter?

Yes - the ddb entries describe the layout of the virtual disk. Snapshots always match the base disk geometry, so they don't need (and couldn't use) extra ddb entries.

My suggestion would be to create a new 20 GB sparse disk and modify the newly created .vmdk header file.

0 Kudos