Hi Chris,
Have you checked if the broken RTPProxy server has enabled the timeout
notification feature (-n parameter)?
Regards,
Razvan Crainea
OpenSIPS Developer
On 14.07.2011 14:28, Chris Martineau wrote:
Hi Razvan,
Further to this I have been monitoring the servers to try and see
which server is causing the issue. It only seems to be one instance on
one server that is causing the issue.
When monitoring the UDP traffic to that instance I see U and L records
that do not contain any codec component and this coincides with the
error seen in opensips. No P request is made because I assume that for
some reason this instance of the rtpproxy in opensips has been marked
somehow. It is not the SIP messages causing this as it only happens on
calls that use this instance.
I thought I could just disable this instance in the nathelper cache
screen of opensips-cp but doing this causes the entire system to stop
processing calls until I re-enble the instance again! This never used
to happen and I do not see any errors in the opensips log when the
system stops.
Overall there seems to be some issues with the interface to rtpproxy
in 1.7.
Help...
Regards
Chris
Hi Razvan,
All the rtpproxy instances are pulled from the latest git and they
were all working okay.
All the instances do work when you first start them but for some
reason it starts throwing this error. It can last quite a while before
this happens so it seems to support playback when you start it but at
some time during running it stops? No errors seem to be thrown at the
rtpproxy end although I only have logging at the ERR level.
What determines if the rtpproxy supports media playback or not? What
could change that would cause it to throw this error mid way through
running. It may be something local but I have no idea where to start
looking seeing as nothing has changed on the media servers between 1.6
and 1.7!
The broken pipe issue seems to only kick in when the call to notify of
timeout occurs. Of the 8 instances it happens on all of them except
the first one started which would point to some sort of clash or error
at the opensips receiving end because it is across both servers. i.e
server1: works,broken,broken,broken server2: broken,broken,broken,broken.
Any help you could offer would be greatly appreciated.
Regards
Chris
*From:*users-boun...@lists.opensips.org
[mailto:users-boun...@lists.opensips.org] *On Behalf Of *Razvan Crainea
*Sent:* 13 July 2011 16:12
*To:* users@lists.opensips.org
*Subject:* Re: [OpenSIPS-Users] 1.7 and rtpproxy issues
Hi Chris,
The error is pretty obvious: one (or more) of your RTPProxy node
doesn't support media playback.
Regarding the RTPProxy broken pipe errors, how often are you getting
those errors?
Regards,
Razvan Crainea
OpenSIPS Developer
On 13.07.2011 18:02, Chris Martineau wrote:
Hi,
Since loading the latest 1.7 I have had a problem with the playback on
rtpproxy.
I startup opensips with 8 rtpproxy instances 4 each on 2 other servers
and all works okay.
After a random period opensips starts throwing an error as follows
ERROR:rtpproxy:rtpproxy_stream: required functionality is not
supported by the version of the rtpproxy running on the selected node.
Please upgrade the rtpproxy and try again.
This seems to happen on one of the servers which gives no playback
while the other seems to work fine playing back as normal.
The only errors I see on rtpproxy are Broken pipe errors which I have
raised previously but haven't had a response yet.
Restart the instances on the effected server and it starts working
again. The servers were built at the same time with the same git
version of rtpproxy and were working fine.
Any ideas?
Many thanks
Chris
_______________________________________________
Users mailing list
Users@lists.opensips.org <mailto:Users@lists.opensips.org>
http://lists.opensips.org/cgi-bin/mailman/listinfo/users
_______________________________________________
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users
