Truenas 13.0 smb constantly uses 100% of a thread

[Deko22]

Hello,
I updated my Truenas 12.0-U8.1 to Truenas 13.0 a few days ago. I have also upgraded my pools via console.
Since then, Samba crashes all the time and it often requires multiple restarts of the machine until I am finally able to access the file I desire.
Every time the connection from my Win10 Pc to a network share freezes, smb starts to use 100% of a cpu thread.
This can only be stopped by restarting the machine and the whole cycle repeats. I then get the following warning message:

​

Code:

The following system core files were found: smbd.core. Please create a ticket at https://jira.ixsystems.com/ and attach the relevant core files along with a system debug. Once the core files have been archived and attached to the ticket, they may be removed by running the following command in shell: 'rm /var/db/system/cores/*'.

I am running a Xeon E5-2667 v2 on a Supermicro X9sri-f combined with 256gb ddr3 1866 registred ram and had the system running for years without problems prior to the update.

Is anyone experiencing the same issue as I do or knows what is going on?
Thanks in advance.

 

[Samuel Tai]

You may be hitting this bug @anodos described here: https://www.truenas.com/community/t...when-accessing-a-directory.101246/post-697141.

However, that was on SCALE. I don't know what the equivalent for Core would be. Can you look at /var/log/samba4/log.smbd, and find any log messages at the same time as your freeze?

 

[anodos]

You may be hitting this bug @anodos described here: https://www.truenas.com/community/t...when-accessing-a-directory.101246/post-697141.

However, that was on SCALE. I don't know what the equivalent for Core would be. Can you look at /var/log/samba4/log.smbd, and find any log messages at the same time as your freeze?

13.0 and SCALE share common code base. There are some users hitting a bug in recycle bin. I had to significantly refactor for samba 4.15 because of troubles with nested datasets and path handling, and some users are hitting bugs in error handling for cases where the connected user has no rights to share.

 

[profwalken]

Hi Anodos
I've just installed a freshTN13 and nextcloud and it seems weird things occurs with content update from TN SMB share to nextcloud.
In fact most of the time folders and files on SMB side are not in real time updated, or not updated at all, in NC. I found a way to force nextcloud side to be updated, checking a box on folders and then select download in NC menu, then I receive a zip file with DL content (this is no use for me, but after this when you go and click on root folder on NC then content is up to date same as SMB side.
This is the only way I found for now to get NC surely updated.
I don't know if this weird behavior can be related to SMB problems you're speaking about, if so do you have an idea if a fix is coming soon ?