Getfacl hangs.

[wswanson]

TrueNAS-12.-U8.1
Disk pools: /mnt/pool01
Dataset: ds1 (pool01/ds1
SMB Share: /mnt/pool01/ds1
Servername \\TRUENAS

I had a sudden power outage, and my server went down hard. Fortunately, it's a pretty passive box, nothing's been written to it in months. It's used as a media server, and even at that, no one's even read from it. Kodi decided to go bonkers, and I was looking at another media player software that would run reasonably well on the Android TV. The power came back up, and when I rebooted, I was able to scan the pool for errors, and bring it back on line. (don't ask me to recall the command used, I found it here. zpool -F or something like that.)

I can see the \\TRUENAS server in network neighborhood, and I can see the \ds1 share, but it fails to map a drive, or if at maps a drive, I can't browse it.

While poking around and trying to troubleshoot this, I've noticed that there's a "ROOT cannot be an SMB user." new rule. While trying to verify *that*, I ran into a problem:

Editing the share's Filesystem ACL or running getfacl causes either the GUI or the console to hang. There's only one share and 3 users, so I can't see that this is some huge datamining exercise for the system.

Rather than trying to un-FUBAR that, I thought it might be easier to simply re-do the share, somehow, and have some confidence that the table/database is now uncorrupted.

I haven't done anything yet!

How do you go about deleting and recreating the share, but keeping the file system (and files) in place? I can even delete all of the users (there's only 3, I think) and the groups (3 made while trying to figure things out.)

 

[anodos]

getfacl hanging is symptomatic of a larger problem (assuming you are invoking it correctly).
The correct syntax is getfacl /mnt/tank/foo. If that hangs, try getfacl -n /mnt/tank/foo

 

[anodos]

If it's still hanging when invoked properly, I'd start looking carefully at your zpool integrity. zpool status check for errors, then run a scrub and check for errors.

 

[wswanson]

Thanks for your time. The web console window is being difficult about what it actually selects. (to get a better chance at capturing the output, I tried to SSH in, and it just doesn't like PuTTY for some reason: can't seem to get the Diffie Hellman negotiation.)

Code:

root@truenas[~]# getfacl /mnt/pool01/ds1
# file: /mnt/pool01/ds1
# owner: root
# group: wheel
user::rwx
group::r-x
other::r-x

root@truenas[~]# zpool status
  pool: boot-pool
 state: ONLINE
  scan: scrub repaired 0B in 00:01:03 with 0 errors on Sat Sep 17 03:46:03 2022
config:
        NAME        STATE     READ WRITE CKSUM
        boot-pool   ONLINE       0     0     0
          da0p2     ONLINE       0     0     0


errors: No known data errors


  pool: pool01
 state: ONLINE
  scan: scrub repaired 0B in 03:30:08 with 0 errors on Sat Sep 10 19:18:12 2022
config:
        NAME                                          STATE     READ WRITE CKSUM
        pool01                                        ONLINE       0     0     0
          gptid/cfabc9a0-77d6-11eb-ba1f-000c29b2db62  ONLINE       0     0     0


errors: No known data errors

I'm running zpool scrub pool01. The last time finished with no errors.

The console still hangs (spins indefinitely) when I go to sharing / SMB and click "Edit Filesystem ACL". I can't reproduce the command line getfacl hang at the moment.

Back to the larger issue: Can the SMB data be reset, and the whole SMB configuration started from scratch, without loss of the files, or is that intimately part of the process? I think I've configured it wrong, or something went sideways when the NAS shutdown unexpectedly, and I can no longer access the files via windows, but I can see both the server and the share.