Hi brian, I think I agree, we found this post ( http://www.mail-archive.com/[email protected]/msg03093.html) which mentions that failure route and branch route will be called on a failure. I suspect that mech underneath engage_media_proxy is modifying SDP c= in failure and branch. I may switch to using use/end mediaproxy, but if that is the case, I figure it may be easy enough to fix in the module itself. So I will look there first and see if I can't make up a patch.
Charles On Fri, Feb 10, 2012 at 4:33 PM, Brian ipt <[email protected]> wrote: > > > On Fri, Feb 10, 2012 at 10:01 PM, Charles Solar <[email protected]>wrote: > >> Definitely not - I know there was a lengthy discussion about calling >> use_media_proxy twice and that function a while ago. The only thing I do >> with a natted invite is this: >> >> # NAT test >> if( is_method("INVITE") && !has_totag() ) >> { >> $avp(can_uri) = $ru; >> >> setflag(3); # flag for NAT >> force_rport(); >> fix_contact(); >> append_hf("P-hint: NAT\r\n"); >> xlog( "L_INFO", "Engaging media proxy" ); >> engage_media_proxy(); >> } >> >> Charles >> >> > Sorry not to be more helpful and you as you say in your mail it does look > like a bug, engage_media_proxy is using dialog and when you failover > mediaproxy module is modifying SDP incorrectly for the second time.. > > use use_media_proxy and end_media_sessions to get around this for the > moment. > > > >> >> On Fri, Feb 10, 2012 at 3:54 PM, Brian ipt <[email protected]> wrote: >> >>> >>> >>> On Fri, Feb 10, 2012 at 9:21 PM, Charles Solar >>> <[email protected]>wrote: >>> >>>> I have been using engage_media_proxy in my script, but it seems to >>>> cause problems when I failover in the dynamic routing module. >>>> This is the scenario: >>>> >>>> 1. Incoming call >>>> 2. Engage media proxy >>>> 3. Try first GW (which is down) >>>> 4. Failure route >>>> 5. Try next GW >>>> 6. Next GW sends Session progress & invite >>>> 7. Media proxy module 'fixes' my SDP contact twice >>>> >>>> My SIP message end up with duplicated SDP contact info. For example:c=IN >>>> IP4 33.66.22.1133.66.22.11 >>>> >>>> If the first gateway works, this does not happen. In fact it works >>>> perfectly. But I guess engage_media_proxy or the dialog module is having >>>> an issue with a failure scenario? >>>> >>>> My failure route is pretty simple: >>>> >>>> failure_route[4] { >>>> xlog("L_INFO", "Failure Route: $rm, $fu, $ru\n"); >>>> >>>> if (use_next_gw()) { >>>> xlog("L_INFO", "Trying next gateway" ); >>>> t_relay(); >>>> exit; >>>> } else { >>>> xlog( "L_INFO", "Exhausted gateway list" ); >>>> t_reply ("503", "Service not available"); >>>> exit; >>>> } >>>> } >>>> >>>> Should I switch to use_media_proxy/end_media_session if I want to >>>> expect failovers? Or could I be doing something wrong? >>>> >>>> Thanks guys >>>> Charles >>>> >>> >>> >>> >>> Hi Charles, >>> >>> Are you using fix_nated_sdp anywhere in your scripts or using anything >>> else that modifies the SDP? >>> >>> Brian >>> >>> >>> >>>> >>>> _______________________________________________ >>>> Users mailing list >>>> [email protected] >>>> http://lists.opensips.org/cgi-bin/mailman/listinfo/users >>>> >>>> >>> >>> _______________________________________________ >>> Users mailing list >>> [email protected] >>> http://lists.opensips.org/cgi-bin/mailman/listinfo/users >>> >>> >> >> _______________________________________________ >> Users mailing list >> [email protected] >> http://lists.opensips.org/cgi-bin/mailman/listinfo/users >> >> > > _______________________________________________ > Users mailing list > [email protected] > http://lists.opensips.org/cgi-bin/mailman/listinfo/users > >
_______________________________________________ Users mailing list [email protected] http://lists.opensips.org/cgi-bin/mailman/listinfo/users
