Hi Saúl, On 4/22/11 4:26 AM, "Saúl Ibarra Corretgé" <[email protected]> wrote:
>Hi Jeff, > >On 04/21/2011 11:14 PM, Jeff Pyle wrote: >> I went over my config and I think the logic is sound. I'll verify it on >> a dev box tonight once traffic is low. >> >> The call flow is serial forking from one carrier gateway to another when >> the first sends a 503. >> >> * load prefs for first gateway >> * use_media_proxy() >> * t_relay() INVITE to first gateway >> * receive 503 >> * end_media_session() in failure_route[] >> * load carrier prefs for second gateway >> * use_media_proxy() >> * t_relay() INVITE to second gateway >> * receive 400 (because of jacked up c= line) >> * end_media_session() >> >> The dispatcher log seems to jive with this. I've got a "remove" command >> from the end_media_session() after the first gateway, and the subsequent >> stats (all 0s), and a "remove" command from the second >> end_media_session() and more stats. I think the commands are hitting >> correctly. >> >> Is it not correct to use(), end(), use(), end()? Once I use(), is it >> impossible to end()?* >> > >Depends on how you do that. OpenSIPS (1.x.x) does not modify the >original message it appends a lump, which is a modification indicator >that will be made when the message is going to be sent out. > >That said, if you call use(), end(), use() *before* sending the message >out you'll be in trouble. So as long as I t_relay() the message between the first use() and the end(), or the end() and the second use(), all will be well? - Jeff _______________________________________________ Users mailing list [email protected] http://lists.opensips.org/cgi-bin/mailman/listinfo/users
