User CIFS Mounts Suddenly Not Working

English Network/Internet
1

Greetings Tumbleweed Community,

I have been on travel for a bit and didn’t use my machine for the past 3 weeks. Upon return, I updated my TW system and my user CIFS mounts are now not working and I am unsure why. I have details below. Any recommendations are appreciated. Thank you! I looked and did not find other forum entries that aligned with this; my apologies if I missed something.

[Message from Terminal when Entering]

> cd /slash-johndoe/ 
bash: cd: /slash-johndoe/: No such device

[Relevant fstab]

//nas/slash-smb                            /slash-johndoe         cifs   vers=default,sec=ntlmssp,credentials=/root/mnt/smb-johndoe.cred,uid=1000,gid=8888,dir_mode=0750,file_mode=0750,x-systemd.automount,x-systemd.mount-timeout=10  0  0

[Relevant /etc/hosts]

192.168.1.11  nas

[Relevant credentials]

> sudo cat /root/mnt/smb-johndoe.cred 
[sudo] password for root:  
username=johndoe
password=nope 
domain=local

[Relevant dmesg output]

> dmesg | grep -i cifs 
  164.823334] Key type **cifs**.spnego registered 
  164.823342] Key type **cifs**.idmap registered 
  164.823932] **CIFS**: Attempting to mount \
as\slash-smb 
  164.862548] **CIFS**: VFS: **cifs**_mount failed w/return code = -22 
  164.867897] **CIFS**: Attempting to mount \
as\slash-smb 
  164.903734] **CIFS**: VFS: **cifs**_mount failed w/return code = -22

[Me Trying to Get More Error Information]

> > sudo umount /slash-johndoe  
> sudo mount --verbose /slash-johndoe/ 
mount.cifs kernel mount options: ip=192.168.1.11,unc=\
as\slash-smb,vers=default,sec=ntlmssp,dir_mode=0750,
file_mode=0750,uid=1000,gid=4200,user=johndoe,domain=local,pass=**** 
mount error(22): Invalid argument 
Refer to the mount.cifs(8) manual page (e.g. man mount.cifs) and kernel log messages (dmesg)

Kind regards,
iqgrande

2

Reads like this thread:

https://forums.opensuse.org/showthread.php/572342-CIFS-mount-fails-with-kernel-5-18-9-1-default?highlight=cifs_mount+-22

See if this helps…

https://forums.opensuse.org/showthread.php/572342-CIFS-mount-fails-with-kernel-5-18-9-1-default?p=3140852#post3140852

Specifically…

Something to try: Does specifying “vers=2.0” or “vers=3.02” in the mount command work for you?

3

Yes, both vers options appeared to have addressed this. Is this a kernel regression that will be fixed eventually (so I can go back to vers=default eventually)? That’s how I interpreted those threads, but I want to make sure I didn’t misread anything. Thanks again and my apologies for missing that thread. I put my question under “Network/Internet” whereas the solution was under “Install/Boot/Login”… I can see the justification for putting it under both. Should these sorts of questions fall under the former or the latter (just so I make sure I am looking/posting in the right location)?

4

All good. Glad to have been of guidance.

I put my question under “Network/Internet” whereas the solution was under “Install/Boot/Login”… I can see the justification for putting it under both. Should these sorts of questions fall under the former or the latter (just so I make sure I am looking/posting in the right location)?

Not a network related issue…the “Install/Boot/Login” probably best given that it is for

Questions about installation, login, boot issues, partitioning, file systems, software that runs at boot (GRUB, LILO, boot scripts)

5

Yes, I’d expect once the fix is in place, using “vers=default” should work reliably as it did previously.

The manpage (man mount.cifs) states…

default - Tries to negotiate the highest SMB2+ version supported by both the client and server.

6

You obviously know the maximum version supported by //nas/slash-smb. Thus the most robust option is specifying this number instead of negotiating.

Putting everything into one line of /etc/fstab is a handy feature. Use it if it works. But troubleshooting and maintenance can be cumbersome. For utmost robustness and ease of maintenance consider the use of local services defined in /etc/systemd These take precedence over vendor settings.

For illustration purposes I converted the following mount to a systemd unit:

//fritz.box/FRITZ.NAS                      /fritz.box              cifs   noauto,username=ftpuser,vers=1.0  0  0

[FONT=monospace]**erlangen:~ #** systemctl cat fritz.box.mount                   
**# /etc/systemd/system/fritz.box.mount**
[Mount] 
What=//fritz.box/FRITZ.NAS 
Where=/fritz.box 
Type=cifs 
Options=credentials=/root/fritz.box.credentials,vers=1.0 
**erlangen:~ #**[/FONT]

Adding automount capability is straight forward:

**erlangen:~ #** systemctl cat fritz.box.automount  
**# /etc/systemd/system/fritz.box.automount**
[Automount]  
Where=/fritz.box 
TimeoutIdleSec=1min 
**erlangen:~ #**

To my experience local services readily survive hundreds of distribution upgrades and are less prone to failure caused by sloppy development.

7

Total nonsense, and not relevant to the reported issue in any way.