Ah, I see. It looks like there was a misunderstanding there. The transport config that you specified is only used to create the SIP (udp) transport, but not the media transports. Since the application doesn't specify the transport config for the media transports, the media transports were created using the default configuration (which retrieves the IP address from the hostname). To override the published address of the media transports, you need to initialize another pjsua_media_config, and call pjsua_media_transports_create() before pjsua_start(). cheers, -benny tloginbr-pjsip at yahoo.com.br wrote: > I'm sending my sample code. just to be clear: My real > IP address is 10.100.0.143, I put in my /etc/hosts > file the ip 10.100.0.199, and in the application I set > the public_addr to 10.100.0.14. Bellow is a piece of > the log output (level 4): > > 14:54:06.944 udp0x83dc09c SIP UDP transport started, > published address is 10.100.0.14:5090 > 14:54:06.944 pjsua_media.c RTP socket reachable at > 10.100.0.199:4000 > 14:54:06.944 pjsua_media.c RTCP socket reachable at > 10.100.0.199:4001 > 14:54:06.945 pjsua_media.c RTP socket reachable at > 10.100.0.199:4002 > 14:54:06.945 pjsua_media.c RTCP socket reachable at > 10.100.0.199:4003 > 14:54:06.945 pjsua_media.c RTP socket reachable at > 10.100.0.199:4004 > 14:54:06.945 pjsua_media.c RTCP socket reachable at > 10.100.0.199:4005 > 14:54:06.945 pjsua_media.c RTP socket reachable at > 10.100.0.199:4006 > 14:54:06.945 pjsua_media.c RTCP socket reachable at > 10.100.0.199:4007 > 14:54:06.945 pjsua_acc.c Account > <sip:10.100.0.14:5090;transport=UDP> added with id 0 > > I'm using the api version 0.8.0 from the site (not > svn) and I compiled it with the options for > PJ_CONFIG_MAXIMUM_SPEED inside config_site.h. > > thanks again, > > Thiago >> tloginbr-pjsip at yahoo.com.br wrote: >>> Thanks a lot. I didn't enable stun, so maybe there >> is >>> a little problem here... But I also didn't disable >> it, >>> so if it comes enabled by default thats how it is. >> In that case, it should work. I've checked the >> source code and I >> couldn't find any problem with it. Could you post a >> simple sample to >> reproduce it? >> >> cheers, >> -benny
