- Annoyance
There is something about this pre-existing zpool that makes unriad unable to create new shares on it from the GUI. I am easily able to create new datasets from the command line and then unraid creates a share (it improperly assigns the dataset generated share primary storage to the array) with the dataset name. Previously in another bug report a user mentioned that behavior (wrong storage assignment) when manually creating datasets from CLI - so that behavior is expected since that isn't the recommended method.
ZFS attributes pasted below and error from syslog when attempting to create a share. I did notice that a few attributes do not match my other newly created post-6.12 zpool that adding shares to works fine on.
root@CVG02:~# zfs get all snapshot
NAME PROPERTY VALUE SOURCE
snapshot type filesystem -
snapshot creation Fri Nov 11 2:47 2022 -
snapshot used 12.8T -
snapshot available 1.18T -
snapshot referenced 41.5K -
snapshot compressratio 1.02x -
snapshot mounted yes -
snapshot quota none default
snapshot reservation none default
snapshot recordsize 128K default
snapshot mountpoint /mnt/snapshot local
snapshot sharenfs off default
snapshot checksum on default
snapshot compression off local
snapshot atime off local
snapshot devices on default
snapshot exec on default
snapshot setuid on default
snapshot readonly off default
snapshot zoned off default
snapshot snapdir hidden default
snapshot aclmode discard default
snapshot aclinherit restricted default
snapshot createtxg 1 -
snapshot canmount on default
snapshot xattr on default
snapshot copies 1 default
snapshot version 5 -
snapshot utf8only off -
snapshot normalization none -
snapshot casesensitivity sensitive -
snapshot vscan off default
snapshot nbmand off default
snapshot sharesmb off default
snapshot refquota none default
snapshot refreservation none default
snapshot guid 8916430419615625548 -
snapshot primarycache all default
snapshot secondarycache all default
snapshot usedbysnapshots 0B -
snapshot usedbydataset 41.5K -
snapshot usedbychildren 12.8T -
snapshot usedbyrefreservation 0B -
snapshot logbias latency default
snapshot objsetid 54 -
snapshot dedup off local
snapshot mlslabel none default
snapshot sync standard default
snapshot dnodesize legacy default
snapshot refcompressratio 1.00x -
snapshot written 41.5K -
snapshot logicalused 13.1T -
snapshot logicalreferenced 13.5K -
snapshot volmode default default
snapshot filesystem_limit none default
snapshot snapshot_limit none default
snapshot filesystem_count none default
snapshot snapshot_count none default
snapshot snapdev hidden default
snapshot acltype off default
snapshot context none default
snapshot fscontext none default
snapshot defcontext none default
snapshot rootcontext none default
snapshot relatime off default
snapshot redundant_metadata all default
snapshot overlay on default
snapshot encryption off default
snapshot keylocation none default
snapshot keyformat none default
snapshot pbkdf2iters 0 default
snapshot special_small_blocks 0 default
root@CVG02:~# zpool status snapshot
pool: snapshot
state: ONLINE
scan: scrub canceled on Sat Jul 1 10:11:14 2023
config:
NAME STATE READ WRITE CKSUM
snapshot ONLINE 0 0 0
raidz1-0 ONLINE 0 0 0
sdk ONLINE 0 0 0
sdi ONLINE 0 0 0
sdj ONLINE 0 0 0
sdg ONLINE 0 0 0
sdl ONLINE 0 0 0
sdf ONLINE 0 0 0
errors: No known data errors
Yes, the scrub was cancelled on purpose because I didn't realize I had it as a scheduled cron task AND unraid can now run it on ZFS pools on a schedule. It had just ran a few days prior. I just pasted that so you could see what kind of pool it is - made out of 10yo 3tb Hitachis.
Error in syslog when trying to create a share on the zpool named "snapshot".
Jul 3 08:27:24 CVG02 shfs: share cache full
Jul 3 08:27:24 CVG02 emhttpd: error: shfs_mk_share, 6451: No space left on device (28): ioctl: /newshare
Jul 3 08:27:24 CVG02 emhttpd: shcmd (405): rm '/boot/config/shares/newshare.cfg'
Jul 3 08:27:24 CVG02 emhttpd: Starting services...
Jul 3 08:27:25 CVG02 emhttpd: shcmd (408): /etc/rc.d/rc.samba restart
This error makes it sounds like it cannot identify the FS properly as it runs shfs_mk_share rather than zfs create.
For comparison - successful share creation on the other newly create zpool:
Jul 3 08:33:56 CVG02 shfs: /usr/sbin/zfs create 'cache/newsharenewpool'
Jul 3 08:33:56 CVG02 emhttpd: Starting services...
Jul 3 08:33:56 CVG02 emhttpd: shcmd (421): chmod 0777 '/mnt/user/newsharenewpool'
Jul 3 08:33:56 CVG02 emhttpd: shcmd (422): chown 'nobody':'users' '/mnt/user/newsharenewpool'
Jul 3 08:33:56 CVG02 emhttpd: shcmd (423): /etc/rc.d/rc.samba restart
Diagnostics attached for posterity. In summary I'm trying to figure out why unraid can't create the datasets on this particular zpool, and if it can be resolved.