[SOLVED] Dir does not mount correctly

[Circlenaut]

Issue:
Backup directory is pointed to the local file-system instead of a file-system on another zfs block device within the same server.

Background:

My /etc/pve/storage.cfg is configured as such:

dir: local
path /var/lib/vz
content vztmpl,iso
shared 0
zfspool: local-zfs
pool rpool/data
content images,rootdir
sparse 1
zfspool: storage
pool storage
content images,rootdir
sparse 1
zfspool: machines
pool machines
content rootdir,images
sparse 1
dir: Backups
path /storage
content backup
maxfiles 10
shared 0

Notice how it references back to /storage.

ZFS says it's mounted to /storage.

zfs get mountpoint storage
NAME PROPERTY VALUE SOURCE
storage mountpoint /storage default

But I don't see it with df

df -h
Filesystem Size Used Avail Use% Mounted on
udev 32G 0 32G 0% /dev
tmpfs 6.3G 9.8M 6.3G 1% /run
rpool/ROOT/pve-1 54G 43G 12G 80% /
tmpfs 32G 43M 32G 1% /dev/shm
tmpfs 5.0M 0 5.0M 0% /run/lock
tmpfs 32G 0 32G 0% /sys/fs/cgroup
machines 23G 0 23G 0% /machines
rpool 12G 0 12G 0% /rpool
rpool/ROOT 12G 0 12G 0% /rpool/ROOT
rpool/data 12G 0 12G 0% /rpool/data
/dev/fuse 30M 60K 30M 1% /etc/pve
tmpfs 6.3G 0 6.3G 0% /run/user/1000

Which is why I think under Datacenter-->Server-->Backups "Backups" shows the same usage level as "local".

How can I get proxmox to recognize and mount Backups under my much larger "storage" zfs pool instead of "local"?

This used to work properly when I first installed v5.1 and up until maybe 3 months ago.

 

[wolfgang]

Hi,

I guess your zfs can't mount because the dir /storage are not empty.

please check your logs and if so you will find an entry or call

mount -all

 

[Circlenaut]

That's it! the dir was not empty. For some reason it was occupied with the backup directory (dump). I supposed this may have happened if /storage did not mount on boot and a backup was started; filling the local dir and thus making it not empty on the next remount attempt.

I did have to remove my back dir from datacenter --> storage. Otherwise the directory "dump" kept being recreated. I did try first "rm -r /storage/dump && mount -all" to try and remount before the service gets a chance to recreated the directory but even when I deleted the directory in the webui, mount -all by itself did not work. I needed to restart the node completely.

Might be a good idea to run a check to see if the mount-point "/storage" exists before the service creates any directories in the mount-point directory.

Thanks!