Today I decided to remove a disk from a vm to attach it to another vm.
The disk file was created with vm (333) and I attached it it to vm (109) without moving the files from the 333 directory to109 directory.
I simply cut the text string from 333.conf and aded it as a new disk to 109.conf and started 109. Bingo, there was the disk.
I then deleted VM 333 (purge)
Disk still present in 109. Abit later I need to reboot the host in order to install a RMS-200 for my ZFS pool upon which all my vm's reside.
It took a peculiarly long time for the host to shut down and I could not figure out why.
Well... PM was busy deleteting that 3TB file!
When I tried booting 109 afterwards it was unable to boot because of some insane feature in proxmox that patiently waits for you to shut down the server so that it can delete a file that is in use on another vm.... It would not start 109 because "blabla disk 333/storage... etc..." does not exist.
The result was, PM decided to wait patiently for 109 to release lock on the 3TB file and then delete it...
So I lost 2.5 TB of data that is absolutely unrecovereable unless I spend 400 USD on klennex zfs recovery tool... and even then, recovery is questionable. Unfortunately, no snapshots!
Seriously... fix this!
Current situation from PM perspective:
-Ohhh... the user has asked me to delete a file! Unfortunately, he attached it to a different VM... so Ill just wait for him to shut down that VM and then delete it! I am sure he does not suspect that to happen! EVILGRIN...
What situation should be like:
-Ohhh... the user has asked to delete a file! Unfortunately, the file is in use by a different VM... Maybe I should just leave it there and not delete it when he shuts down that vm!... I am sure he gets mad if I just delete something that seems to be needed! I am such a good boy.
The disk file was created with vm (333) and I attached it it to vm (109) without moving the files from the 333 directory to109 directory.
I simply cut the text string from 333.conf and aded it as a new disk to 109.conf and started 109. Bingo, there was the disk.
I then deleted VM 333 (purge)
Disk still present in 109. Abit later I need to reboot the host in order to install a RMS-200 for my ZFS pool upon which all my vm's reside.
It took a peculiarly long time for the host to shut down and I could not figure out why.
Well... PM was busy deleteting that 3TB file!
When I tried booting 109 afterwards it was unable to boot because of some insane feature in proxmox that patiently waits for you to shut down the server so that it can delete a file that is in use on another vm.... It would not start 109 because "blabla disk 333/storage... etc..." does not exist.
The result was, PM decided to wait patiently for 109 to release lock on the 3TB file and then delete it...
So I lost 2.5 TB of data that is absolutely unrecovereable unless I spend 400 USD on klennex zfs recovery tool... and even then, recovery is questionable. Unfortunately, no snapshots!
Seriously... fix this!
Current situation from PM perspective:
-Ohhh... the user has asked me to delete a file! Unfortunately, he attached it to a different VM... so Ill just wait for him to shut down that VM and then delete it! I am sure he does not suspect that to happen! EVILGRIN...
What situation should be like:
-Ohhh... the user has asked to delete a file! Unfortunately, the file is in use by a different VM... Maybe I should just leave it there and not delete it when he shuts down that vm!... I am sure he gets mad if I just delete something that seems to be needed! I am such a good boy.
Last edited:
