Unable to mount thin-provisioned LVM volume?

Jan 12, 2015
93
2
28
Hi,
I have six Proxmox hypervisors (various licenses) running 5.4 which I began upgrading this week (to 6.2-4). I'm on the second machine today and running into the same problem that I had on my first upgrade.

There is an LVM volume defined from the 5.4 Proxmox install which is tagged with thin-provisioning flags. It will not mount after the upgrade. It was mounting OK on 5.4 but booting 6.2 after the upgrade just goes into emergency mode. I've double-checked my other 5.4 hosts currently in production and they are using this same type of partition without any issues so it's not something changed during the 6.2 upgrade.

I've managed to work around it by adding the noauto flag to the data volume in /etc/fstab on my 6.2 hosts. This allows the host to boot normally but the partition must be mounted by hand after logging into the machine.

# lvs
LV VG Attr LSize Pool Origin Data% Meta% Move Log Cpy%Sync Convert
danger pass_thru -wi-ao---- 931.00g
data pve twi-a-tz-- 1.67t 0.00 0.15
root pve -wi-ao---- 96.00g
swap pve -wi-ao---- 8.00g


What is the proper method for mounting the /dev/pve/data partition?
 
Last edited:

aaron

Proxmox Staff Member
Staff member
Jun 3, 2019
1,341
156
63
How were those boxes installed? With the PVE installer or on top of Debian?

The `data` thin provisioned volume is not meant to be mounted directly but is used to store the VM disks on a block level.

The /etc/fstab of a fresh install should look like this:
Code:
$ cat /etc/fstab
# <file system> <mount point> <type> <options> <dump> <pass>
/dev/pve/root / ext4 errors=remount-ro 0 1
/dev/pve/swap none swap sw 0 0
proc /proc proc defaults 0 0

Code:
$ lvs
  LV            VG  Attr       LSize  Pool Origin Data%  Meta%  Move Log Cpy%Sync Convert
  data          pve twi-aotz-- 28.37g             7.34   1.68                           
  root          pve -wi-ao---- 14.75g                                                   
  swap          pve -wi-ao----  7.00g                                                   
  vm-100-disk-0 pve Vwi-a-tz--  2.00g data        5.20                                   
  vm-100-disk-1 pve Vwi-a-tz--  1.00g data        4.78                                   
  vm-100-disk-2 pve Vwi-a-tz--  1.00g data        4.78
Note how the block device volumes for the VMs use `data` as pool.
 
Jan 12, 2015
93
2
28
Thanks for the info. The partitions on these Proxmox hosts have been through at least three, maybe four, upgrades so far. We've always used the PVE installer (even for upgrades) so the partitions have always been defined by the installer. I'm not sure how 'data' got into /etc/fstab but it's been that way for as long as I can remember. It could be something I "fixed" a long time ago not totally understanding how thin provisioning works and it's been carried through all our upgrades. I think we tried using block devices for VMs right from the start but ended up putting them on mounted SAN partitions instead so they could easily be brought up on different proxmox hosts.
 

About

The Proxmox community has been around for many years and offers help and support for Proxmox VE and Proxmox Mail Gateway. We think our community is one of the best thanks to people like you!

Get your subscription!

The Proxmox team works very hard to make sure you are running the best software and getting stable updates and security enhancements, as well as quick enterprise support. Tens of thousands of happy customers have a Proxmox subscription. Get your own in 60 seconds.

Buy now!