Cannot properly resize LXC volume

[petrij98]

For the past couple of months, I have been desperately trying to resize the volume of my Proxmox container because it has been constantly running out of space. I followed the few guides that I've seen on the forum page, but every single one of them has led me to a dead end. When I utilize the GUI to resize my filesystem, instead of properly doing so and adding space to root, it adds it into /dev/shm and that directory refuses to stay unmounted. I am terrified to manually shrink and grow the filesystem without proper guidance, as I have already experienced firsthand those consequences on my file server. Any help on this issue would be greatly appreciated.

This is one of the main errors I encounter when attempting to run a filesystem check:



No forum post I have seen as been able to resolve the "Is a directory while trying to open" error. This is absolutely mounted already and is properly loaded whenever I turn the container on.

When I run a df -h within the container, this is what I'm greeted with:



My theory is that it may be as a result of accidentally added the extra space while the container is still running, but again, any ideas would be incredibly helpful.

 

[Gilou]

Are you trying to use e2fsck on a BTRFS subvolume?

Can you provide the result of

findmnt /

?

If what I assume is correct, you'll need to check how to resize a btrfs partition, not ext4. Something along the lines of https://btrfs.readthedocs.io/en/latest/btrfs-filesystem.html#man-filesystem-resize

 

[petrij98]

Here you go!



It's reporting as ext4, so I'm not sure if trying to resize btrfs will do me any good.

 

[Gilou]

I'm not sure what you're trying to do now.. So your first screenshot was inside the container, not on PVE right?

So … can you do findmnt / inside the same container?

Also, could you show us that container configuration (pct config 105 most likely), and especially the rootfs property?
And… then, the storage configuration for that rootfs storage (something like pvesm status -storage xxx xxx being the rootfs)

 

[petrij98]

None of my screenshots were taken from inside the container. They were all from Proxmox. This is what it looks like from within the container:

Not sure if it being a zfs file type would make a difference, but that's what it populates as from within the container.

rootfs within my config file says the following: "rootfs: VM-Import:subvol-105-disk-0,size=202G"

And below is my pvesm status for the drive that rootfs is hosted on:

 

[Gilou]

OK. It is rather unclear what you want. I'm guessing.. you want to resize the ZFS storage used for all your containers, right? The zfspool called VM-Import? If so, you need to have space available on the underlying device…

Or is it the container volume itself?

 

[petrij98]

I only want to resize the file system for one container. The ZFSPool is called VM-Import, which is where the bootdisk is hosted. There was space on the underlying device. I increased it from 103G to 202G via the GUI. According to the container Summary in the GUI, the bootdisk is still stuck at 103G.

 

[Gilou]

Do you see the disk as having the good size in Proxmox GUI? (or using pvesm list VM-Import | grep ' 105$' ),
What does the subvolume tell you, from the host? (zfs list testzfs/subvol-105-disk-0)

It's indeed not normal for the container to not have been resized by the GUI process, it might be a bug, or there might be an issue with your zpool…