ESXi 3.5 Disk Full - delete all snapshots

[dualboot]

Hi All,

I have a virtual machine running FreeNAS with 5 x 200G disks defined as a JBOD. I've been filling it up with backup data in preparation for some system rebuilding, when it hung with the 'no more space for snapshot file (thats the jist of it - can't recall the exact error message now.

I couldn't understand where all my space had gone, but have been on a steep learning curve, and now understand about delta files. I have got into the file system via SSH and can see that there are delta files taking up hundreds of Gigs. Snapshot manager showed no snapshots (nothing to delete). I saw a recommendation to take a new snapshot and then do a 'delete all'. I have done this, and the physical server's hard drive light has been on now for two hours, solid.

Is this to be expected. I don't want to be hasty and interrupt something that is working as it should, but is there any way I can confirm it is working and not just in some infinite loop ?

One supplemental question. To stop this from happining again, can I just edit the config of the VM and make these disks independent ? Can I do that for an existing virtual machine without loosing the data on the disk ?

[java_cat33 · Accepted solution]

Is there only one snapshot in place? Definitely don't lose your patience and try and stop the committing process.

You can watch the snapshot deletion/committ process by the following command (will need to login with unsupported console access - http://www.virtualizationadmin.com/articles-tutorials/vmware-esx-articles/general/how-to-access-the-...

watch "ls -Ghtu --full-time *.vmdk" -d

In regards to your question regarding stopping this from happening again - yes you can make the disk independent persistent. This must be done while the VM is powered off. You wont lose data making this change either.

Message was edited by: java_cat33 - Added not about unsupported console access for ESXi

[dualboot]

Thanks Java_Cat -

I kept my cool, and left it overnight. Looking at the timestamps this morning the first disk took 3 hours to consolidate. There were 5 x 200G disks so it took all night. But now FreeNAS is running again, and my mammoth rsync is running again. (Glad I did an rsync not cp now ! )

Thanks again for the help

[java_cat33]

Good work - there's plenty of free scripts on the web that you can schedule to run against your environment to identify snapshots etc. It's something I'd definitely recommend. Most health check scripts these days will report on active snapshots which can save you a lot of head aches....