Hi Halomoan,

loose_route() only marks the Route hdr to be deleted when the message is sent out (it does not actually delete the hdr). So, the search after loose_route() should work, but it look a bit broken to me, as it will check all the Route hdrs and not only the one belonging to the server.

regards,
bogdan

Halomoan Chow wrote:
I copy and paste the logic example generated from sipwise below.
I can see for first request, it marked the sender is behind nat with record_route(";nat=yes")

In the loose route block, it tries to identify sender if they are behind nat with search("^Route:.*;nat=yes").

In one sip proxy architecture, after you enter loose_route() function, the Via with ;nat has gone, right? and the search("^Route:.*;nat=yes") test would always return false.
Please correct me if i'm wrong. Thank you.


        if(!is_method("REGISTER"))
        {
                if(nat_uac_test("19"))
                {
                        record_route(";nat=yes");
                }
                else
                {
                        record_route();
                }
        }
        if(is_method("CANCEL") || is_method("BYE"))
        {
                end_media_session();
        }

        if(loose_route())
        {
                if(!has_totag())
                {

xlog("L_INFO", "Initial loose-routing rejected - M=$rm RURI=$ru F=$fu T=$tu IP=$si ID=$ci\n"); sl_send_reply("403", "Initial Loose-Routing Rejected");
                        exit;
                }
                if(nat_uac_test("19") || search("^Route:.*;nat=yes"))
                {
                        fix_nated_contact();
                        setbflag(6);
                }

                route(3);
        }


On 9/12/07, *Bogdan-Andrei Iancu* <[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>> wrote:

    Hi Halomoan,

    Mainly because the NAT detection on SIP replied is very poor - you can
    check only if the contact address is a private one or not. The
    other nat
    test are not applicable for replies (but only for requests).

    Anthow, when you process the request, you learn that the calee is
    behind
    NAT, so you can simple remember this (in RR) than detecting it
    each time
    when receiving a reply.

    regards,
    bogdan

    Halomoan Chow wrote:
    > Dear All
    >
    > I don't understand why in the Openser+MediaProxy i need
    > record_route_preset("x.x.x.x:5060;nat=yes"); and
    > search("^Route:.*;nat=yes")
    > What is the purpose while I can do client_nat_test to see if the
    > request is come from a nat-ed client.
    >
    > If I remove search("^Route:.*;nat=yes") from below statement, I
    can do
    > re-INVITE without any problem because it still have
    client_nat_test to
    > find out the nat in between.
    >
    > if (client_nat_test("3") || search("^Route:.*;nat=yes")) {
    >    setbflag(6);
    >    use_media_proxy();
    > };
    >
    >
    > Hope someone can give me a clue.
    >
    > Thank you.
    >
    > Regards,
    >
    > Halomoan
    >
    ------------------------------------------------------------------------
    >
    > _______________________________________________
    > Users mailing list
    > Users@openser.org <mailto:Users@openser.org>
    > http://openser.org/cgi-bin/mailman/listinfo/users
    >




_______________________________________________
Users mailing list
Users@openser.org
http://openser.org/cgi-bin/mailman/listinfo/users

Reply via email to