Smbd exited in signal 6 since last update ( FreeNAS-9.10.2-U4)

[anodos]

Disabling oplocks is not feasible for me; since nobody's complained this far, I'll just deal with the daily kernel errors on my mailbox.

That's mostly the conclusion I reached as well. I'm seeing the same on my servers.

 

[Cellobita]

Again, thank you very much for your help and info. Take care!

 

[Cellobita]

Well, today one of my customers lost connection to their FreeNAS box, approximately at the same time the server logged the core dump error

 

[anodos]

Well, today one of my customers lost connection to their FreeNAS box, approximately at the same time the server logged the core dump error

Well, there are several open bug reports for the problem on bugs.freenas.org. I'm not sure how to yell much louder. @Ericloewe any thoughts?

Perhaps you can contribute to one of the tickets stating that this is affecting your clients.

Otherwise, you may need to downgrade to U3 and apply some mitigations for the recent samba CVE. :(

 

[Cellobita]

Never mind, my post was informational only, just to let other folks know that it may cause disruption to their users; this was the first time over the past few days that this issue rose from a mere annoyance to something requiring attention.

I am grateful for your help with this, though - hope the bug gets squashed soon!

 

[Cellobita]

I did follow up both the FreeNAS bug report at https://bugs.freenas.org/issues/24225 and the Samba one at https://bugzilla.samba.org/show_bug.cgi?id=12798 - it does seem to be the issue I'm experiencing. Also, the mention to corrupted files has left me worried - I'm doing a rollback tonight on all U4 boxes, until we have a fix.

 

[anodos]

I did follow up both the FreeNAS bug report at https://bugs.freenas.org/issues/24225 and the Samba one at https://bugzilla.samba.org/show_bug.cgi?id=12798 - it does seem to be the issue I'm experiencing. Also, the mention to corrupted files has left me worried - I'm doing a rollback tonight on all U4 boxes, until we have a fix.

You may need to
(1) set the auxiliary parameter nt pipe support = no on the servers (nasty fix).
Or
(2) change the datasets you're sharing to have the "exec = off" property zfs set exec = off Tank/Share

I still have to test if (2) is an effective mitigation for the recent Samba CVE (I'm pretty sure it is). I still need to set up a testing environment for it.

 

[Cellobita]

Will do. Thanks!

 

[anodos]

Will do. Thanks!

I think I deleted a part of my original response. The above parameters are if you need to rollback to U3. U4 fixed a particularly nasty remote code execution vulnerability.

 

[Ericloewe]

I'm not sure how to yell much louder. @Ericloewe any thoughts?

They seem to be on it now. Apparently, iX in TN is in the process of moving across town and that might be introducing some delays in the dev process (that AD ticket has also been holding things up forever, delaying RC4).

Probably a good idea to ping the ticket tomorrow afternoon, anyway.

 

[Cellobita]

I think I deleted a part of my original response. The above parameters are if you need to rollback to U3. U4 fixed a particularly nasty remote code execution vulnerability.

No problem, you had mentioned that before. Much appreciated.

 

[themac]

I have the same problem on a client's site since one week after updating to FreeNAS-9.10.2-U4.
But it's a big problem there since they use MS Access and it's constantly saying network connection lost and then the user needs to force shut down Access and reconnect.

 

[anodos]

I have the same problem on a client's site since one week after updating to FreeNAS-9.10.2-U4.
But it's a big problem there since they use MS Access and it's constantly saying network connection lost and then the user needs to force shut down Access and reconnect.

9.10.2-U5 is coming soon. https://bugs.freenas.org/projects/freenas/issues?query_id=392

 

[themac]

9.10.2-U5 is coming soon. https://bugs.freenas.org/projects/freenas/issues?query_id=392

Thanks

Do you have any idea when the update will be released? The issue is causing quite some trouble and I want to know if it's worth downgrading or just waiting a little longer.

 

[Dice]

..I just got the same error messages in the daily security output. Figured I'd chime in, adding to the stock of affected users.
Running version: FreeNAS-9.10.2-U4 (27ae72978)

Lagoona.local kernel log messages:
> pid 20843 (smbd), uid 0: exited on signal 6 (core dumped)
> pid 25414 (smbd), uid 0: exited on signal 6 (core dumped)
> pid 29322 (php), uid 80: exited on signal 11
> pid 26058 (smbd), uid 0: exited on signal 6 (core dumped)
> pid 36032 (smbd), uid 0: exited on signal 6 (core dumped)
> pid 40543 (smbd), uid 0: exited on signal 6 (core dumped)
edit:
However, perhaps worth mentioning: I use (don't recall reasoning to why) Server maximum protocol SMB3_11. (min: ------)
And the regular Auxillary parameters:

ea support = no
store dos attributes = no
map archive = no
map hidden = no
map readonly = no
map system = no

 

[anodos]

Thanks

Do you have any idea when the update will be released? The issue is causing quite some trouble and I want to know if it's worth downgrading or just waiting a little longer.

The update has been released. http://download.freenas.org/9.10/STABLE/9.10.2-U5/

 

[Nicholas M]

Ive just had something similar happen on FreeNAS-11.0-U1 (aa82cc58d)..

Code:

Jul 20 12:22:35 abney kernel: Failed to fully fault in a core file segment at VA 0x800f64000 with size 0x3000 to be written at offset 0x2a1000 for process smbd
Jul 20 12:22:35 abney kernel: Failed to fully fault in a core file segment at VA 0x800f64000 with size 0x3000 to be written at offset 0x2a1000 for process smbd
Jul 20 12:22:35 abney kernel: pid 96668 (smbd), uid 0: exited on signal 6 (core dumped)

(Ive posted this in FreeNAS > Development> Bug reporting Discussion > incase we need to raise a new bug report.)