[SR-Users] Issue with Socket selection on Forwarding ACK message

[Emmanuel BUU]

Hi Timmo,

We do have *enable_double_rr* set to 1 but the route set added by the 
proxy when record_route() on the INVITE consists in a single route. This 
is because Alice and Bob are BOTH registered with the alternate port 
5066 on the same interfacec and using the same protocol (UDP). To us, 
this seems logical.

Do you have a double route in your case even though both party are on 
the same port, same interface and both UDP? If so, could you send us an 
exemple of Route: header?

Emmanuel BUU
IVèS

Le 2018-02-02 à 14:18, Timmo Verlaan a écrit :
> Hi Thomas,
>
> We have a similar situation but we use double route headers so the 
> correct egress socket is chosen. You can enable this by setting a 
> modparam: enable_double_rr.
> Would this be a solution for you?
>
> Kind regards,
> Timmo Verlaan
>
> On 2 Feb 2018 13:57, "Thomas Carvello" <thomas.carvello at ives.fr 
> <mailto:thomas.carvello at ives.fr>> wrote:
>
>     Thank you for you answer.
>
>     We have tried to change the local port for Bob, but it doesnt
>     change anything. And the contact value in 200 OK message has no
>     influence in this case.
>
>     In fact, we have made a further investigation regarding the socket
>     selection *and read the code. *The issue seems to be located in
>     the RR module and the loose_route() function.
>
>     In the  after_loose() function in loose.c, the  function
>     set_force_socket()  is called only if a DOUBLE route is mentioned
>     in the route set of the ACK message
>
>     But when both users are using 5066 as proxy port, we get only ONE
>     route for the proxy in the route set (and to us it is OK). In this
>     case, we get a trace:
>
>     "No next URI found"
>
>     and the code exits. Later in the message processing, when
>     t_relay() is called, the forward_request() selects the first
>     socket defined in our configuration instead.
>
>     At this point, we can't presume what socket we be select. We
>     believe that it is a software bug  and that after_loose() should
>     force the send_socket even though we have only one route in the
>     route set. We also checked the 5.1 code and there is no change in
>     this module that would alter this behavior.
>
>     Are we missing something?
>
>     Thank you for you time,
>
>     Thomas
>
>     Le 26/01/2018 à 15:43, Евгений Голей a écrit :
>>     Hi
>>
>>     Could it be because of Bob happend to use 5060 as local port?
>>     Yes, the port and the address in the ACK are indicated by what
>>     the value in Contact was in reply 200 Ok. Look at the message 200 Ok
>>
>>
>>         Четверг, 25 января 2018, 13:00 +03:00 от Thomas Carvello
>>         <thomas.carvello at ives.fr> <mailto:thomas.carvello at ives.fr>:
>>
>>
>>         Hello,
>>
>>         i have an issue with my Kamailio 4.1.9 configuration.
>>
>>         This configuration is multi-homed, we have*two network*
>>         interfaces, one on a private network and on the public
>>         Internet. Kamailio is configured to listen on port 5060 and
>>         5066 on both interfaces. We register two users Alice and Bob
>>         on the  public Internet using port 5066.  Both users are
>>         behind a NAT and we capture the SIP exchange on the proxy server.
>>
>>         We have set the parameter mhomed=1
>>
>>         When Alice calls Bob, we have
>>
>>         Alice			Proxy				Bob
>>
>>         src=5063		dst=5066
>>         INVITE ------------------>
>>
>>                                  src=5066
>>                                  ------  INVITE ---------------> dst=5060
>>
>>                                  dst=5066
>>                                  <------- 200 OK -------------- src=5060
>>
>>
>>         dst=5063
>>         <------- 200 OK --------- src=5066
>>
>>         src=5063		dst=5066
>>         -------- ACK ----------->
>>
>>                                  *src=5060 (blocked by NAT)*
>>                                  ------  ACK-----x            dst=5060
>>
>>
>>         The ACK packet gets relayed with the wrong source port. Then
>>         the NAT rejects the packet and the call cannot be established.
>>
>>         For some reason, when Bob calls Alice, the call is correctly
>>         established. Could it be because Bob happend to use 5060 as
>>         local port?
>>
>>         Also, if we set nhomed=0 it works BUT we are not sure that
>>         multi homed is handled correctly.
>>
>>         I was wondering if you have encounter this issue before?
>>
>>         I have investigated the code for selection socket and what is
>>         the logic of this selection ?
>>
>>         /*How does kamailo knows that it should choose 5066 as src
>>         port if the user is registered using port 5066 instead of 5066?*
>>         /
>>
>>         Thank you for your time.
>>
>>         Thomas
>>
>>
>>
>>         _______________________________________________
>>         Kamailio (SER) - Users Mailing List
>>         sr-users at lists.kamailio.org <mailto:sr-users at lists.kamailio.org>
>>         https://lists.kamailio.org/cgi-bin/mailman/listinfo/sr-users
>>         <https://lists.kamailio.org/cgi-bin/mailman/listinfo/sr-users>
>>
>>
>>
>>     Best
>>     Evgeniy
>
>
>     _______________________________________________
>     Kamailio (SER) - Users Mailing List
>     sr-users at lists.kamailio.org <mailto:sr-users at lists.kamailio.org>
>     https://lists.kamailio.org/cgi-bin/mailman/listinfo/sr-users
>     <https://lists.kamailio.org/cgi-bin/mailman/listinfo/sr-users>
>
>
>
> _______________________________________________
> Kamailio (SER) - Users Mailing List
> sr-users at lists.kamailio.org
> https://lists.kamailio.org/cgi-bin/mailman/listinfo/sr-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.kamailio.org/pipermail/sr-users/attachments/20180202/593a5155/attachment.html>