Bug 7522 - s3 member server winbind becomes unresponsive when used with s4
s3 member server winbind becomes unresponsive when used with s4
Status: RESOLVED INVALID
Product: Samba 4.0
Classification: Unclassified
Component: AD: LDB/DSDB/SAMDB
unspecified
Other Linux
: P3 normal
: ---
Assigned To: Andrew Bartlett
samba4-qa@samba.org
:
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2010-06-17 06:57 UTC by Lukasz Zalewski
Modified: 2011-01-21 05:22 UTC (History)
1 user (show)

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Lukasz Zalewski 2010-06-17 06:57:17 UTC
once winbind is started on s3 member server (3.3.8) it occupies 100% of CPU and becomes unresposive (can't be pinged using wbinfo -p or queried)

s4 logs gets filled up with multitude of the following repeats:
added interface ip=my.s4.ip.address nmask=255.255.254.0
cldap netlogon query domain=my.s4.domain host=MYS3MEMBER user=(null) version=6 guid=(null)

Let me knw if you want me to provide additional log dumps or tcptrace

Thx

Luk
Comment 1 Matthias Dieter Wallnöfer 2010-06-17 07:25:19 UTC
Well, I CC metze on this bug since he knows much about winbind on s3 and probably also something about CLDAP in s4.
Comment 2 Stefan Metzmacher 2010-06-17 07:52:00 UTC
a tcpdump would be a good start
see http://wiki.samba.org/index.php/Capture_Packets

If possible please also try if it's already fixed in 3.3.13
and or 3.5.3.

Are you using the current master branch on the dc?
Comment 3 Matthias Dieter Wallnöfer 2010-06-18 12:35:43 UTC
Do you know if the CLDAP server returns NT_STATUS_OK?
Comment 4 Lukasz Zalewski 2010-06-18 15:39:47 UTC
Hi Matthias,
After each 
cldap netlogon query domain=my.s4.domain host=MYS3MEMBER user=(null) version=6
guid=(null)
entry in s4 samba log there is a huge list of ldb_trace_request(s)
each one ending on

ldb: ldb_trace_response: DONE
error: 0

After multiple blocks of the above
added interface ip=my.s4.ip.address nmask=255.255.254.0
appears and the whole history repeats itself

I do not see any NT_STATUS_OK there. Which specific log files should i be looking at?

Thx

Luk
Comment 5 Matthias Dieter Wallnöfer 2010-10-31 12:54:01 UTC
Is this still an issue?
Comment 6 Lukasz Zalewski 2010-10-31 14:07:20 UTC
Hi Matthias,
I have not tested this for some time and probably won't till Christmas (we will bring our build to the current master around then). I can test it tomorrow but it will be with quite old (over a month old) git revision. 
Comment 7 Matthias Dieter Wallnöfer 2011-01-18 14:50:33 UTC
How far are we here?
Comment 8 Lukasz Zalewski 2011-01-20 16:37:52 UTC
Hi Matthias,
I'm shamed to say that due to the limited time during Christmas vac, but mainly due to the fact that our current s4 setup is working so well, we decided to postpone the upgrades till Easter vac. I will need to tie some loose ends in the next few months so wont have any time to look into it. Maybe we should close this bug and I open a new one if the issue happens after the upgrade, unless you are happy to leave it open and we re-visit it in the next few months

Regards

Luk
Comment 9 Matthias Dieter Wallnöfer 2011-01-21 05:22:23 UTC
We close it for now with "INVALID" and you reopen it if you are able to reproduce the issue.
I prefer to close bugs with unknown status to have more overview about the other ones. But if the problem turns out to still persist then I encourage to reopen.

(In reply to comment #8)
> Hi Matthias,
> I'm shamed to say that due to the limited time during Christmas vac, but mainly
> due to the fact that our current s4 setup is working so well, we decided to
> postpone the upgrades till Easter vac. I will need to tie some loose ends in
> the next few months so wont have any time to look into it. Maybe we should
> close this bug and I open a new one if the issue happens after the upgrade,
> unless you are happy to leave it open and we re-visit it in the next few months
> 
> Regards
> 
> Luk
>