Hi all,

 

I have a raidz2 ZFS pool with 8 main drives and an optane drive that is split in 2 partitions, one assigned as SLOG device and the other as L2ARC. The pool was created under Unraid 6.11 using the ZFS plugin.

 

Since upgrading to 6.12 (and the removal of the ZFS plugin), I've been manually mounting the ZFS pool at array start with a user script. Everything was working perfectly.

 

I'm currently on 6.12.15 and I've decided to import the pool in Unraid to benefit from the extra UI integration.

However, with the pool imported in Unraid and having disabled the script that manually mounts the pool, none of the datasets in the pool get mounted and the status in the bottom left on the Unraid page is stuck at Array starting - Mounting disks...

If instead I manually mount the pool, it works correctly.

 

I've attached the diagnostics before manually mounting the pool

Also, you can find below the output of zfs list

root@Jarvis:~# zfs list
NAME              USED  AVAIL     REFER  MOUNTPOINT
disk1             641G   282G      128K  /mnt/disk1
disk1/appdata    9.53G   282G     9.53G  /mnt/disk1/appdata
disk1/data       11.1G   282G     11.1G  /mnt/disk1/data
disk1/docker     8.37G   282G     8.37G  /mnt/disk1/docker
disk1/dockers      96K   282G       96K  /mnt/disk1/dockers
disk1/domains     589G   282G      589G  /mnt/disk1/domains
disk1/isos       22.7G   282G     22.7G  /mnt/disk1/isos
disk1/libvirt     110M   282G      110M  /mnt/disk1/libvirt
disk1/system     38.2M   282G     38.2M  /mnt/disk1/system
zfs              8.70T  11.9T      273K  /mnt/zfs
zfs/Backup       1.28T  11.9T     1.28T  /mnt/zfs/Backup
zfs/Dati         1.85T  11.9T     1.22T  /mnt/zfs/Dati
zfs/Download     18.6G  11.9T     18.6G  /mnt/zfs/Download
zfs/Film         4.11T  11.9T     4.11T  /mnt/zfs/Film
zfs/Foto          687G  11.9T      673G  /mnt/zfs/Foto
zfs/Musica       5.00G  11.9T     4.99G  /mnt/zfs/Musica
zfs/TimeMachine   677G  11.9T      520G  /mnt/zfs/TimeMachine
zfs/test          110G  11.9T      110G  /mnt/zfs/test
zfs/test2        39.0M  11.9T     39.0M  /mnt/zfs/test2
zfs/test4        5.55G  11.9T     5.55G  /mnt/zfs/test4
zfs/test5        39.0M  11.9T     39.0M  /mnt/zfs/test5

 

disk1 is the main Unraid array which is also formatted as ZFS and doesn't have any issue with mounting the datasets. 

The problematic pool is zfs

 

Below you can find the output of zfs mount

 

root@Jarvis:~# zfs mount
disk1                           /mnt/disk1
zfs                             /mnt/zfs

 

I found this bug report that was describing the same issue that I'm encountering but in that case the culprit was the sharenfs feature being enabled which in my case is off for all datasets.

Quote

 

 

Any suggestion on how to get the datasets to mount correctly?

 

jarvis-diagnostics-20250128-2147.zip

Edited January 28, 20251 yr by Andrea3000

11 hours ago, Andrea3000 said:

optane drive that is split in 2 partitions, one assigned as SLOG device and the other as L2ARC.

This is not supported by the GUI, only one partition per device.

 

 

2 hours ago, JorgeB said:

This is not supported by the GUI, only one partition per device.

 

 

Thanks for the reply.

Is there any plan to support multiple partitions in the GUI?

29 minutes ago, Andrea3000 said:

Is there any plan to support multiple partitions in the GUI?

AFAIK not in the near future.