[OpenSIPS-Users] from address in the t_relay

Bogdan-Andrei Iancu bogdan at opensips.org
Fri May 18 06:51:49 EDT 2018


Hi Ryan,

To answer your question on t_realy() and how it decides on the sending 
interface - normally OpenSIPS tried to preserve the interface (outbound 
<= inbound) if the protocol and routing allows it (if you do not 
explicitly change proto or interface). If preserving is not possible 
(incoming via UDP, but destination is TCP), it will use the first 
listener matching the needed outgoing protocol.
Definitely the SIP FROM is not involved at all in this logic !!

Regards,

Bogdan-Andrei Iancu

OpenSIPS Founder and Developer
   http://www.opensips-solutions.com
OpenSIPS Summit 2018
   http://www.opensips.org/events/Summit-2018Amsterdam

On 04/18/2018 03:59 PM, Esty, Ryan wrote:
>
> Hi list,
>
> I’m hoping someone can direct me to some documentation or help me out. 
> I’m trying to send an invite to something outside my domain. 
> Everything seems to work but no packets come back to opensips. I see 
> the packets on the wire coming to the machine but it looks like the 
> tcp port that was used to send stuff out isn’t there anymore. A 
> colleage found this gem in the log: “proto_tcp_send: Successfully 
> connected from interface 209.197.207.36:5060 to 209.197.207.36:5060” 
> thinking that maybe the tcp connection was opened by mistake with the 
> local side not set correctly.
>
> My question is when t_relay is called how does it know the from 
> address? I printed out my from domain $fd and it is using my internal 
> domain, which works out to be a 192.168 address. The address listed 
> above is from a successfully DNS look up on the request URI domain and 
> it works out to be the above address.
>
> Ryan
>
>
>
> _______________________________________________
> Users mailing list
> Users at lists.opensips.org
> http://lists.opensips.org/cgi-bin/mailman/listinfo/users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensips.org/pipermail/users/attachments/20180518/90c900d0/attachment-0001.html>


More information about the Users mailing list