Vmware Consolidation Failed

Consolidation often fails. Don't use it!

see my EE Article on snapshots, and understand them.

HOW TO: VMware Snapshots :- Be Patient

and one thing to learn with snapshots is

1. Be Patient.

You have 40GB free, and three snapshots, 40GB is dangerously low, did you try to consolidate when the VM was on?

can you confirm the size of the virtual disk, 1 and 2.

I would handle it this way....power off the VM.

1. Take a new snapshot of the VM.
2. Wait 60 seconds.
3. in snapshot manager, select DELETE ALL.
4. Wait and be patient....
5. it can seem to hang stop at 90, 95 or 99%

No, I did not try to consolidate while the vm was on.  Should I?
Not sure what you mean by confirm the size of the disks.  The vm has 3 disks.  Disk 1 is this disk in question with the snapshots and it's thick provisioned for 500GB.  Disk 2 and 3 are in a different datastore on a different device.  I powered on the vm and verified that the C: drive is 500GB.

No do not try consolidation whilst the VM is on!

Okay, just wanted you to check, because....you have three snapshots, and this VM has been running on snapshots for a while, and the performance of the VM will be poor on snapshots...

snapshot 1 is 300GB approx
snapshot 2 is 100GB approx
snapshot 3 is 300GB approx

700GB of snapshot changes to merge into the parent, this could take a while to merge....

and with only 40GB free....

that is not alot of working over head, for this to work....

can you free up the datastore, or do you have another datastore we can work with....

because an easy method is to use the vCenter Server CLONE option, right click the VM > CLONE, select a new datastore with space for the two virtual disks....

and the CLONE process will merge the snapshots into the parent, and then after power up this new VM, check working, and delete the old.

OR, you can use the clone tool (vmkfstools), at the console or SSH remotely, just to clone a single disk, based on the last snapshot, let me know, and I'll let you know the command to use.

Thanks!  So, in the process of seeing what space I could free up on other datastores, I've found a few other vms running on snapshots.  I cloned DC01 which is a smaller vm to a different datastore and will move some others around after I verify everything is working properly after the clone and can delete the original.  Once I move these "slimmed down" vms, I might be able to get enough room to move EX01.  Just in case I can't get enough space, please let me know the command you'd recommend to clone the individual disks.  Thanks again.

You will still need space, to CLONE, and you use the vmkfstool -i

I have 4 datastores to work with but none are big enough to clone the entire VM.  I have the following free:
Datastore1:  79.74GB free
Datastore2: 532.86GB free
Datastore3: 223.16GB free
Datastore4:  199.80GB free

So can I clone the individual disks to whatever datastore they'll fit in, reconfigure the vm to use those disks, delete the old disks after everything is up and running properly, and then clone the whole vm back into Datastore4?
I've never used the vmkfstool thing before.

Awesome!  Thanks for you help on this.
Should I do it via ssh straight to the host or the vCenter?

ssh remotely to the host.

So our file that contains the local password for the host is out of commission right now.  Is there a way to reset the root password through vCenter?  I'm using domain creds for vCenter.  It's 5.1.

So our file that contains the local password for the host is out of commission right now.  Is there a way to reset the root password through vCenter?  I'm using domain creds for vCenter.  It's 5.1.

I'm afraid not.

You will need to have the root password, to ssh or login via the console.

Connectwise was finally able to repair the configuration file that had the pw.  I'll finish this this week and update.  thank you

Hi,
Sorry for the delay.  Here is what I did:
1. cd /vmfs/volumes/55116ed4-ff2e4cd0-6468-90b11c2b7b58
2. The folder inside is stupidly named with a space in it so I did cd "r2d2 r2d2-cl-EX01", right?
3.  It looks like 2 more frickin snapshots have been created since I was last in there so I did:
 vmkfstools -i "R2D2 - R2D2-CL-EX01-000005.vmdk" /vmfs/volumes/5519e3b9-e3d2a2b6-a832-90b11c2b7b58/EX01VMDK/R2D2-CL-EX01.vmdk

Now I just have a > prompt.  Is that right??  If that is right, how will I know when it's done?  I don't see the new folder created in the 5519e datastore yet.

Now that I think about it, I added a Windows Storage Server on this POS DX4000 a couple days ago and tried to clone the vm to it.  It kept failing at 59% when it appeared that the Windows Storage Server datastore would magically go grey and seem to disconnect.  Odd that it happened at 59% everytime.  I think that's when the 2 additional snapshots were created.

no that does not look correct.

you should not get a > prompt.

you need to create all folders in advance, the vmkfstools command, will expect to be able to write to the folder.

you can use datastore friendly names, to make it easier, rathe than the GUID numbers

I did get a > prompt and it doesn't seem to care what I type at the > prompt--it just goes back to >

the command you are issuing is incorrect

There we go.  I had to create the EX01VMDK folder.  It's showing a Clone: % done now.  Thank you!

That's what I wrote! http:a#41420665

you need to create all folders in advance, the vmkfstools command, will expect to be able to write to the folder.

I know, that's why I said "Thank you".

okay....no problems...good luck.

Just checked on it.  
Says "Clone:  39% done."  
Put me back at the "/vmfs/volumes... # prompt.
In the datastore, there is now a 524GB vmdk file.  
Do you think it finished or what's up with the 39%?

that's a little odd....

is that the size of the parent ? it should be ?

you've not run out of storage space ?

I cloned the snapshot vmdk into a datastore that had room after I moved some stuff around.  There is now 32GB left.  524GB is the size of the original parent in the datastore that contains all the snapshots.

It's complete then, odd, it normally, states 100%

I would always recommend testing the new cloned vmdk, and once you have confirmed, it's exactly the same as the original, delete the original files and snapshot (or safer to move/archive)

It looks to have booted up properly and is running well.  I think we might be all good.  This Exchange server was thick provisioned for 500GB but I've migrated them to O365 hybrid so now there's 433GB free of the 500 provisioned for the C: drive and 142GB free of the 200GB provisioned for the D: drive.  When I finish cloning the snapshot of the D: drive, I can just clone the whole virtual machine using thin provisioning and not use up the space, right?.

I can just clone the whole virtual machine using thin provisioning and not use up the space, right?.

Theory yes, other than Microsoft does not support Exchange on thin disks, and you may find, that all the blocks in the vmdk have changed, and you may not save as much space in the thin disk, unless you prepare the virtual disk before conversion to thin!

but this discussion is another question, and off topic, and not really anything to do with the original question submitted.