VMware Cloud Community
sjscott2003
Contributor
Contributor

ESX 3.5 snapshots, not enough drive space left to commit--Help please

I have ESX 3.5 on a Dell PowerEdge 1850, with a 60gb hard drive. I have two VMs running on this server and have snapshots on one of them that is consuming all but 2 gb of the available drive store. I have a Buffalo Terastation NAS on my network that I could move the smaller less relavent of the two vm's off, but that will only free about 8gig. It appears that I cannot mount a USB External 1tb drive to the host, from my readings so far.

The larger of the two vm's is not running right now due message from server that:

There is no more space for the redo log of server-000005.vmdk. You may be able to continue this session by freeing disk space on the relevant partition, and clicking Retry. Otherwise, click Abort to terminate this session.

Currently this vm is powered off, but I need to get it back online quickly because it is our internet filter.

What is the best way to proceed with:

1) freeing space to commit the snapshots that are consuming the host disk, (use of usb external seems to be out due to not supporting usb on esx host) 2) order of deleting snapshots in the snapshot manager, to make best use of space during the commit process.

Any help/thoughts would be greatly appreciated.

Thanks

Reply
0 Kudos
5 Replies
f10
Expert
Expert

Hi Scott,

I didnt get the amount of provisioned, used and available space that is there to commit snapshots. However to commit snapshots the free space required should be one and half times of the largest snapshot file. You can check that by logging into the ESX console

  1. cd /vmfs/volumes/datastore/VM

  2. ls -ltrh

Here the size of the snapshots disks would be reported in GB/MB, so lets say if the largest snapshot file is 2GB and you have 3 to 3.5GB available space, you are good to go.

Let me know if you have any other questions.

If you found this or other information useful, please consider awarding points for "Correct" or "Helpful".

Regards, Arun Pandey VCP 3,4,5 | VCAP-DCA | NCDA | HPUX-CSA | http://highoncloud.blogspot.in/ If you found this or other information useful, please consider awarding points for "Correct" or "Helpful".
Reply
0 Kudos
Troy_Clavell
Immortal
Immortal

...or use VMware Converter to move the guest in question to a LUN with more space This will commit the snapshot(s)

Reply
0 Kudos
a_p_
Leadership
Leadership

2) order of deleting snapshots in the snapshot manager, to make best use of space during the commit process.

If you didn't try a "Delete All" yet and the snapshot manager shows the snapshots correctly, then you can commit the snapshots one-by-one, always committing the one closest to the base disk. This way you will not need any additional disk space, because the blocks of the snapshot are written directly to the base disk (which I assume is thick provisioned).

Make sure you wait until one snapshot is committed completely before you start to delete the next snapshot. In case of large snapshot files, the GUI may time out after a few minutes, but the snapshot is still committing. Open a datastore browser and make sure the snapshot file you select is being deleted before starting the next one.

André

Reply
0 Kudos
sjscott2003
Contributor
Contributor

Andre,

Before I understood the snapshot commit process, I did what instinctively made sense at the time, and that was to delete the last snapshot that I made, which is farthest from the base disk, and closest to the your are here. I did not do a delete all. I have a base .vmdk file set to 20g, only 2gb available currently on the system. current vmdk sizes are:

.000001.vmdk at 1.5gb

.000002.vmdk at 3gb

.000003.vmdk at 7.8gb

.000004.vmdk at 5.3gb

.000005.vmdk at 4.3gb

.000006.vmdk at 3.8gb

I am also showing to have:

snapshot1.vmsn 1gb

snapshot2.vmsn 1gb

snapshot3.vmsn 1gb

snapshot4.vmsn 1gb

snapshot6.vmsn 1gb

missing snapshot5.vmsn ? Maybe the one I tried deleting?

Thoughts on proceeding?

Reply
0 Kudos
a_p_
Leadership
Leadership

2GB is not a lot, however it should be enough. To be on the secure side, shut down the other VM. This will free the disk space used for the swap file. Then start with deleting the snapshots in the way I described.

André

Reply
0 Kudos