[OpenSIPS-Users] siptrace picks incorrect source socket

[Jeff Pyle]

Hi Bogdan,

Now it won't start.  I see the following errors on config check:

Feb  6 21:21:03 [30051] ERROR:siptrace:parse_siptrace_uri: Invalid key
<version> in trace id!
Feb  6 21:21:03 [30051] ERROR:siptrace:parse_siptrace_id: invalid uri
<3;transport=udp;>
Feb  6 21:21:03 [30051] ERROR:siptrace:parse_trace_id: failed to parse
siptrace uri [3;transport=udp;]
Feb  6 21:21:03 [30051] CRITICAL:core:yyerror: parse error in config file
/usr/local//etc/opensips/opensips.cfg, line 225, column 20-21: Parameter
<trace_id> not found in module <siptrace> - can't set
Feb  6 21:21:03 [30051] ERROR:core:main: bad config file (1 errors)
Feb  6 21:21:03 [30051] NOTICE:core:main: Exiting....


The script has:

   223 loadmodule "siptrace.so"
   224 modparam("siptrace", "trace_on", 1)
   225 modparam("siptrace", "trace_id", "[tid]uri=hep:127.0.0.1:9060
;version=3;transport=udp;")


This is on 2.3/devel git revision 2bcf891 from around 01:00 UTC Feb 07.



- Jeff


On Sun, Feb 5, 2017 at 11:00 AM, Bogdan-Andrei Iancu <bogdan at opensips.org>
wrote:

> Hi Jeff,
>
> Thank you for detailed report. I was able to reproduce and fix it. Please
> see:
>     https://github.com/OpenSIPS/opensips/commit/
> b30af734cdb84991e1f906e3920a94e87c33ea04
>
> If you confirm the fix, I will do the backporting to 2.2 too.
>
> Thanks and Regards,
>
> Bogdan-Andrei Iancu
> OpenSIPS Founder and Developerhttp://www.opensips-solutions.com
>
> On 02/04/2017 04:41 AM, Jeff Pyle wrote:
>
> Hello,
>
> I recently enabled siptrace on an OpenSIPS 2.2.2 system acting as a
> registrar and a proxy.  It has one IPv4 address with several ports for UDP,
> TCP and TLS.  In a case where the INVITE is relayed from TLS to UDP, the
> replies to the UAC are incorrectly being reported as coming from the UDP
> socket.
>
> In the below scenario the UAC is at 64.65.66.67, and its ephemeral TCP
> client port for this call is 61235.  The OpenSIPS proxy is at
> 131.132.133.134 listening on UDP 5060 and TLS 5061.  Also on
> 131.132.133.134 there is a Freeswitch media server listening on UDP 5080.
> The UAC sends an INVITE to the proxy over TLS which routes it to the media
> server over UDP.  The replies relayed to the UAC are reported as having
> come from port 5060 over UDP when in reality they have come from port 5061
> over TCP (TLS).
>
> My config:
>
> listen=udp:131.132.133.134:5060
> listen=tls:131.132.133.134:5061
> listen=hep_udp:127.0.0.1:9030
>
> loadmodule "siptrace.so"
> modparam("siptrace", "trace_on", 1)
> modparam("siptrace", "trace_id", "[hep]uri=hep:127.0.0.1:9060;
> transport=udp;")
>
>
>
> Debugs:
>
>
> INVITE in from UAC over TLS
> Feb  3 21:20:22 testproxy /usr/sbin/opensips[24673]:
> DBG:siptrace:pipport2su: proto 22, host 64.65.66.67 , port 61235
> Feb  3 21:20:22 testproxy /usr/sbin/opensips[24673]:
> DBG:siptrace:pipport2su: proto 22, host 131.132.133.134 , port 5061
>
> INVITE out to media server over UDP
> Feb  3 21:20:22 testproxy /usr/sbin/opensips[24673]:
> DBG:siptrace:pipport2su: proto 17, host 131.132.133.134 , port 5060
> Feb  3 21:20:22 testproxy /usr/sbin/opensips[24673]:
> DBG:siptrace:pipport2su: proto 17, host 131.132.133.134 , port 5080
>
> 100 Trying in from media server over UDP
> Feb  3 21:20:22 testproxy /usr/sbin/opensips[24650]:
> DBG:siptrace:pipport2su: proto 17, host 131.132.133.134 , port 5080
> Feb  3 21:20:22 testproxy /usr/sbin/opensips[24650]:
> DBG:siptrace:pipport2su: proto 17, host 131.132.133.134 , port 5060
>
> 180 Ringing in from media server over UDP
> Feb  3 21:20:22 testproxy /usr/sbin/opensips[24651]:
> DBG:siptrace:pipport2su: proto 17, host 131.132.133.134 , port 5080
> Feb  3 21:20:22 testproxy /usr/sbin/opensips[24651]:
> DBG:siptrace:pipport2su: proto 17, host 131.132.133.134 , port 5060
>
> 180 Ringing out to UAC over TLS (even though it reports UDP)
> Feb  3 21:20:22 testproxy /usr/sbin/opensips[24651]:
> DBG:siptrace:pipport2su: proto 17, host 131.132.133.134 , port 5060
> Feb  3 21:20:22 testproxy /usr/sbin/opensips[24651]:
> DBG:siptrace:pipport2su: proto 22, host 64.65.66.67 , port 61235
>
> 200 OK in from media server over UDP
> Feb  3 21:20:22 testproxy /usr/sbin/opensips[24651]:
> DBG:siptrace:pipport2su: proto 17, host 131.132.133.134 , port 5080
> Feb  3 21:20:22 testproxy /usr/sbin/opensips[24651]:
> DBG:siptrace:pipport2su: proto 17, host 131.132.133.134 , port 5060
>
> 200 OK out to UAC over TLS (even though it reports udp)
> Feb  3 21:20:22 testproxy /usr/sbin/opensips[24651]:
> DBG:siptrace:pipport2su: proto 17, host 131.132.133.134 , port 5060
> Feb  3 21:20:22 testproxy /usr/sbin/opensips[24651]:
> DBG:siptrace:pipport2su: proto 22, host 64.65.66.67 , port 61235
>
> ACK in from UAC over TLS
> Feb  3 21:20:23 testproxy /usr/sbin/opensips[24673]:
> DBG:siptrace:pipport2su: proto 22, host 64.65.66.67 , port 61235
> Feb  3 21:20:23 testproxy /usr/sbin/opensips[24673]:
> DBG:siptrace:pipport2su: proto 22, host 131.132.133.134 , port 5061
>
> ACK out to media server over UDP
> Feb  3 21:20:23 testproxy /usr/sbin/opensips[24673]:
> DBG:siptrace:pipport2su: proto 17, host 131.132.133.134 , port 5060
> Feb  3 21:20:23 testproxy /usr/sbin/opensips[24673]:
> DBG:siptrace:pipport2su: proto 17, host 131.132.133.134 , port 5080
>
>
> Everything routes properly, but it isn't reported by siptrace properly.
> Is this a bug or am I doing something wrong?
>
>
>
> - Jeff
>
>
>
>
>
> _______________________________________________
> Users mailing listUsers at lists.opensips.orghttp://lists.opensips.org/cgi-bin/mailman/listinfo/users
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensips.org/pipermail/users/attachments/20170206/4ecb3c3c/attachment-0001.html>