Deleting a huge 500GB Snapshot ...datastore run out of space

[baubau01]

Hey guys,

So I deleted a huge 500GB snapshot and i thought I had enought space on the datastore to do it`s thing .... Now I am currently out of space on that datastore... what now ? would the snapshot eventually get removed and clean up or what ?

Really appriciated it..

Thanks.

[baubau01]

wow... insufficiend space...

Now I have another VM with a huge as snapshot see pics..

So it looks like the snapshots is 531.72 GB

and the vm that has the snapshot it yyz-file-1 ...

I have only 603.21 free space

Would I have enought to commit this snapshot to disk ? I know it will take a lot of time to do it...

Thanks.

[a_p_]

There could be different reasons for this. To understand what was causing the datatstore to run out of disk space, I need some information:

• which version/build of ESX(i) are you running?
• how much disk free space do you currently have on the datastore?
• do you have other datastores with free disk space (how much)?
• are the VM's virtual disks thin or thick provisioned?
• do you have multiple snapshots or just one?
• attach the latest vmware.log file from the VM's folder.
• post a screen shot of the datastore browser window, showing all the files (names, sizes, date/time, ...) of the VM

André

[baubau01]

So let`s deal with this second issue as it`s more critical:

I have a VM file server that has 3 vmdk`s in one datastore

40 GN Thick

1.9 TB Thick

900 GB Thick

Which doesn`t make any sense... because a datastore has a limit of 2TB ... and if those were thick.. But what`s what it displays unders the VM Hard disk 1 , 2 and 3.

That datastore has 603 space free and has snapshots of 532.40 GB.. would it be able to commit it ?

[a_p_]

You are right, that does not make much sense/is not possible. So please go through my previous post and provide the necessary information.

André

[baubau01]

PMed you.

For your questions:

• which version/build of ESX(i) are you running?

the VM that has the snapshot is on a ESX 4.1.0 260247. Snapshot was taken by Veeam and never got deleted.

• how much disk free space do you currently have on the datastore?

Well if I look at the VM it looks like it`s sitting on 3 datastores. But if I click on the hdd to see where they stay they all show that they are in sata_1 datastore.....which has 603.2 GB free

• do you have other datastores with free disk space (how much)?

I have others but not too much space free

• are the VM's virtual disks thin or thick provisioned?

they say they are thick but if I look at the datastore they seem thin....

• do you have multiple snapshots or just one?

one large one vmware is reporting under datastore management

• attach the latest vmware.log file from the VM's folder.

will attach it soon.

• post a screen shot of the datastore browser window, showing all the files (names, sizes, date/time, ...) of the VM

Poested

[baubau01]

The orther 2 pics

[a_p_]

The good news first. You should be able to commit/delete the snapshots without issues.

Let me explain:

The reason you see same location for the vmdk's in the VM's settings is because all snapshots are located in the VM's "home" folder on the sata datastore. The base vmdk's of the disks however are located on the other two datastores. The base disks are definitely thin provisioned and I can't tell you why it shows thick!?

Snapshot "yyyz-file-1-000002.vmdk" (~192 GB) belongs to the virtual disk the apps datastore witch has ~1.1 TB free disk space, so no issue with this one. The other two snapshots "yyyz-file-1-000001.vmdk" (~2 GB) and "yyyz-file-1_1-000001.vmdk" (~363 GB) have their base disks on the filer datastore, which currently has ~440 GB free disk space. Unless I'm mistaken and according to the screen shots, you have enough free disk space on the apps and filer datastores to safely commit/delete the snapshots. Just make absolutely sure the backup application (Veeam) does not run during this time and tries to create another snapshot!

André

[baubau01]

wow that`s crazy... thanks a lot for your answer. Will remove the snapshot tmr afterwork as we have a long weekend and hope by the weekend to be removed. I know for 100gb takes about 5..6 hours so for 530Gb must take at least 30 plus hours.

Will get you to the other issue as well with some other snapshot that i tried to delete and instead of deleting one by one i thought delete all could do it.. lazy has to pay now lol as my datastore run out of space... but that`s not so critical as this other one..

Thanks, will keep u posted.

[a_p_]

Before you are going to hit the "Delete" button, please double check the file sizes of the snapshots and the free disk space, to make sure this did not change too much!

André

[DSTAVERT]

I would also check to see what the last date was for a completed backup.

[baubau01]

thanks guys first problem is solved.

the more complex one is here:

I have this to-fs1-2k8 vm that has 2 hard drives residing on 2 datastores. Lots of snapshots there 15 or so and by deleting all it run out of space on the datastore. Need to fix this today asap. Please let me know what`s the best options here. I only have another datastore with 300GB space I could use it temporarly. This is a standalone ESX 4.0 full no vcenter ..no other hosts.

[a_p_]

Which build of ESX do you run? Unless you are at least running Build Number 261974 (Update 2), you may need another 2TB datastore to clone the virtual disk.

Maybe there's a chance to recreate the VMSD file to delete the snapshots one-by-one , therefore I would need the current vmsd file as well as the latest vmware.log and vmx file.

André

[AntonVZhbankov]

Deleting huge snapshots is always a challenge.

Best ways to do it:

1) Power VM off before deleting snapshot

2) Clone VM instead of deleting snapshot