asterisk users - Mar 2006 - MeetMe freezes machine with Junghanns QuadBRI cards

Hello everybody,

I've got Asterisk 1.2.4 running with two Junghanns QuadBRI cards using the 
qozap driver from bristuff 0.3.0-PRE-1l. One of the cards is running in TE 
mode, the other one in NT mode.

Whenever I call into an empty MeetMe conference room on one of the NT ports or 
via SIP and hang up the call during the "you are currently the only 
participant in this conference" greeting, the Asterisk server freezes. I
can
still ping the machine, but all userspace processes, system logging, kernel 
logging etc. are hung. The machine takes about 15 minutes to recover from 
this state. At that point, the system load has risen to about 15-20.

If I call into the server on one of the TE ports while the system is frozen, 
it instantly un-freezes and resumes normal operation.

Does anyone else experience the same problem?

Cheers,
Henning Holtschneider
--
LocaNet oHG - http://www.loca.net
Lindemannstrasse 81, D-44137 Dortmund
tel +49 231 91596-25, fax +49 231 91596-55
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
Url :
http://lists.digium.com/pipermail/asterisk-users/attachments/20060323/4ac0aa0c/attachment.pgp

On 3/23/06, Henning Holtschneider <henning@loca.net>
wrote:
> I've got Asterisk 1.2.4 running with two Junghanns QuadBRI cards using
the
> qozap driver from bristuff 0.3.0-PRE-1l. One of the cards is running in TE
> mode, the other one in NT mode.
Have you (or can you) tried it with 0.3.0-pre1k ?

I will be trying the same tomorrow (I'm using 1 quadBRI) and let you know..

cheers

On Thu, 2006-03-23 at 19:25 +0100, Henning Holtschneider
wrote:
> Hello everybody,
> 
> I've got Asterisk 1.2.4 running with two Junghanns QuadBRI cards using
the
> qozap driver from bristuff 0.3.0-PRE-1l. One of the cards is running in TE 
> mode, the other one in NT mode.
> 
> Whenever I call into an empty MeetMe conference room on one of the NT ports
or
> via SIP and hang up the call during the "you are currently the only 
> participant in this conference" greeting, the Asterisk server freezes.
I can
> still ping the machine, but all userspace processes, system logging, kernel
> logging etc. are hung. The machine takes about 15 minutes to recover from 
> this state. At that point, the system load has risen to about 15-20.
> 
> If I call into the server on one of the TE ports while the system is
frozen,
> it instantly un-freezes and resumes normal operation.
> 
> Does anyone else experience the same problem?
A test showed the same behavior at our Asterisk server. Some debugging
revealed that a tight loop in app_meetme.c:conf_flush() consumes the
whole CPU power (HANGUP/NULL frames are not handled correctly while
discarding frames of the channel). I've attached a very small patch
(just 2 lines added) that fixes this. The same bug is still present in
Asterisk 1.2.5

Regards,
	Mike
-- 
Dr. Michael Neuhauser                phone: +43 1 789 08 49 - 30
Firmix Software GmbH                   fax: +43 1 789 08 49 - 55
Vienna/Austria/Europe                      email: mike@firmix.at
Embedded Linux Development and Services    http://www.firmix.at/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: meetme-hangup-freeze.patch
Type: text/x-patch
Size: 426 bytes
Desc: 
Url :
http://lists.digium.com/pipermail/asterisk-users/attachments/20060323/4ef9b800/meetme-hangup-freeze.bin

On 19:25, Thu 23 Mar 06, Henning Holtschneider wrote:
> Hello everybody,
> 
> I've got Asterisk 1.2.4 running with two Junghanns QuadBRI cards using
the
> qozap driver from bristuff 0.3.0-PRE-1l. One of the cards is running in TE 
> mode, the other one in NT mode.
> 
> Whenever I call into an empty MeetMe conference room on one of the NT ports
or
> via SIP and hang up the call during the "you are currently the only 
> participant in this conference" greeting, the Asterisk server freezes.
I can
> still ping the machine, but all userspace processes, system logging, kernel
> logging etc. are hung. The machine takes about 15 minutes to recover from 
> this state. At that point, the system load has risen to about 15-20.
> 
> If I call into the server on one of the TE ports while the system is
frozen,
> it instantly un-freezes and resumes normal operation.
> 
> Does anyone else experience the same problem?
This issue has been discussed before on the chan-sccp
mailinglist.
It appears there's something going wrong in the hangup call
in asterisk.

This issue has been confirmed with Junghanns, sangoma and
digium hardware.
In our tests, the machine dissapeared from the network for
like 5 minutes, so you are lucky to get ping back ;)

Our fix was to get an IAX did for the meetme, and call that
external DID for meetme. This seems to work great for now.

Sorry I can only confirm this and cannot come with a fix

-- 
Michiel van Baak
michiel@vanbaak.info
http://michiel.vanbaak.info
GnuPG key: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x7E0B9A2D

"Why is it drug addicts and computer afficionados are both called
users?"