There is no more space for virtual disk 'Server_01-000004.vmdk'. You might be able to continue this session by freeing disk space on the relevant volume, and clicking Retry. Click Cancel to terminate this session.

[MicaelO]

This showed up on our ESXi 5.0 (where we only have ONE Virtual Machine) and there's a number of files in the datastore that I suspect are removable.

However, I certainly don't want to risk a complete failure by deleting files that shouldn't be deleted, hence this post.

Here's a listing of the files present and their size in the datastore:

Server_01.nvram, 8,48 KB, modified 5 May 2019

Server_01.vmdk, 262 144 000,00 KB, modified 17 Apr 2015

Server_01.vmsd, 1,44 KB, modified 15 Mar 2019

Server_01.vmx, 2,88 KB, modified 6 May 2019

Server_01.vmxf, 0,26 KB, modified 31 May 2018

Server_01-000001.vmdk, 2 327 552,00 KB, modified 15 May 2015

Server_01-000002.vmdk, 103 892 000,00 KB, modified 31 May 2018

Server_01-000003.vmdk, 61 604 860,00 KB, modified 15 Mar 2019

Server_01-000004.vmdk, 43 697 150,00 KB, modified 15 Mar 2019

Server_01-Snapshot3.vmsn, 27,81 KB, modified 31 May 2018

vmmcores-1.gz, 5 114,64 KB, modified 15 Oct 2015

vmmcores-2.gz, 16 672,00 KB, modified 6 May 2019

vmware.log, 30,71 KB, modified 6 May 2019

vmware-19.log, 80,31 KB, modified 31 May 2018

vmware-20.log, 112,06 KB, modified 1 Oct 2018

vmware-21.log, 81,49 KB, modified 14 Dec 2018

vmware-22.log, 133,43 KB, modified 6 May 2019

vmware-23.log, 31,05 KB, modified 6 May 2019

vmware-24.log, 30,71 KB, modified 6 May 2019

According to the VM Properties Server_01-000004.vmdk is used as the Virtual Disk.

Can I just delete any of the oldest vmdk's or are the somehow depending of each other?

The System Requirements of the disk of the server running as a VM is just 75 GB so the vmdk's seems to be very large, especially if they're dependent of each other... :smileyshocked:

Hope someone can help me with this!

BR,

Micael

[IRIX201110141]

Oh.... maybe a new records holder.... a 4 years forgotten snapshot... not bad 😕

Your VM holds multible snapshots which mean that for a 75GB vDisk you needs up 4 times the space (worst case). Please dont delete anything... the -00000x.vmdk depends on each other.  You need to free up some space and deleting (merge together) the snapshots to their current states.

The problem is that for deleting a snapshot you need additional disk space! Also the process of snapshot deletions was changed over the years and your vSphere 5.0 is around 8 years old.

By any change you can add some disk space with NFS and than we can create a copy of the vDisk by using vmkfstools -i Server_01-000004.vmdk /path/new/Server_01.vmdk which merge all files together into a single 75GB vDisk.

Regards

Joerg

[MicaelO]

Thank you Joerg,

Fells good to be a record holder... Or NOT! :smileygrin:

Oh, okay... I was hoping for a quicker solution.

Then I need to order some disks to put into the server and make another array to be used as a new datastore.

It's strange but if I look at the datastore from the Client it says that 33,68 GB is free out of the total capacity of 460,75 GB, however "Last Update" is 14 Mar 2019...

I wonder if I could ge another route and use an external HDD as a temporary storage?

After merging, (and backing up files) delete the old stuff and go back to the RAID Array?

/Micael

[IRIX201110141]

Here is a explanation about vSphere ESXi Snapshots: VMware Knowledge Base

You have 33GB left?....  You have sayed somehting about 75GB vDisk size but for me it looks like a 250GB one. When this is not a Thinprovisioning vDisk its maybe save to click the delete snapshot button within the vSphere Client. But someone with more experience should answered.

You can delete the 2 vmcore files to get some more space back. Also as long as the VM is shutdown there is no swap file created.

Adding a Datastore trough NFS is for me always the easiest thing because no reboot is needed and also no fancy network changes (depends). After the first vmkfstools you have to use this VMDKs in your VM configuration and find out if the VM boots. If so... you clean up the original files.... and run a vmkfstools command again to copy it back. Again.. we are not moving files.. we are create copies for safety reasons.

Regards

Joerg

[MicaelO]

Ah... Okay, any suggestions of a good unit to use for NFS?
I have never used anything like that except a NAS from Neatgear or two... Have no idea what filesystem they have, though.

Well the required space is said to be 75GB, but I made the vdisk larger since we only where to use the one and only server there.

BR,

Micael

[IRIX201110141]

Acording to Deleting virtual machine snapshots without wasting disk space you dont need extra space as long you vDisk isnt a Thinprovisiong one. So check this out and when its a Lazy or Eager Zerothick one its save to shutdown the VM and press the "Delete All Snaps" button within the Snapthotmanager.

Your Snaps are 40,60 and 100GB so it can take some time. Dont trust the progressbar.... be patient!

Regards

Joerg

[MicaelO]

Thanks!

That looked great until I opened Snapshot Manager...

No matter which Snapshot I select the buttons below are all grayed out.

All I can do is to change the name of the Snapshot... :smileyshocked:

BR,

Micael

[jburen]

If you have found some extra disk space you could try to clone the VM to get rid of the snapshots.

[MicaelO]

Guys! :smileygrin:

I actually found a Synology NAS-unit that has been decommissioned and was just sitting at the office. It has 2x1TB (mirrored).

So I suspect that I need to Add hardware for the VM in it's properties?

What Disk Provisioning should I use? Thin Provision, is that the best? The old datastore is probably Thick Lazy Zeroed (chosen by default in the wizard).

Anything else I should think of?

BR, Micael

[IRIX201110141]

Is the NFS already mounted on the ESX? If not create the Volume/Share on the Synology and set a propper ACL. On a real linux the NFS needs the No_Root_Squash  option which is most likely a different term in the Synology Config. Iam pretty sure you can find a guide when use google with "sysnology nfs setup for ESXi".

I dont like the GUI to mount the NFS and preferer the old "esxcfg-nas" and a 2nd. putty for looking into var/log/vmkernel.log or maybe hostd.log. Because when the mount went fail the GUI often dont get a hint whats going wrong.

Let me now when the NFS Datastore is mountet.

Regards,

Joerg

[MicaelO]

The NFS is already mounted, like so:

This is where I stopped and thought of what I should choose:

I presume that I need to add it as a HDD in the VM?

And then what kind of provisioning should I choose?

Thanks in advance!

BR, Micael

[MicaelO]

Some images didn't get to be visible...? Strange, they get to be double when I paste them..

First IMG:

Last IMG:

[IRIX201110141]

Ok..

1. Create a new VM named "n-central2" without a vDisk and choose Datastore "SynologyNFS"

2. Type the following

cd /vmfs/volumes/datastore1/N-Central
vmkfstools -i N-Central-000004.vmdk ../../SynologyNFS/n-central2/n-central2.vmdk -d thin

3. After the clone is finished add the existing n-central2.vmdk to the new VM n-central2 together with a  LSI Parallel controller

4. Start the VM and open the console to see if the VM is booting fine

If the VM boots we have create a working VM and merge all snapshot together into one.

5. Delete the original N-Central VM

Warning:

- In your Screenshots the VM is named N-Central but in our beginning Posts its named Server_01

- i prefer lower case for Datastore and VM names with no space and fancy special chars so we dont need quoting the file names on command line

If you open a 2nd putty you can use

esxtop
s2
d

and

u

to watch disk speeds durin cloning. The "u" will shown the NFS mount.

Regards,

Joerg

[jburen · Accepted solution]

Or you could just clone the original VM to a new VM that you store on the new NFS datastore. Then shutdown the original VM and start de cloned VM to see if everything is working like it should. When you're happy you can delete the original VM and move the cloned VM to the correct datastore. When you're all done you can dismount the NFS datastore. Personally, I think this is easier than typing commands in a putty session...

[IRIX201110141]

Status?

[MicaelO]

Okay, here's what has happened:

I tried to do the SSH and CLI commands, but I somehow couldn't get the server to start properly.

VMware said it was running but I couldn't reach the GUI of the server...

So I went for the copy solution and it worked.

So now I can use it again, will do a clean install on the previous vDisk and start from there.

Thanks for both of your efforts! Very appreciative over here!