Hello, On git master from today, I have a simple sbc configuration:
outbound_interface=lan next_hop=10.9.8.7:5060 next_hop_for_replies=yes enable_rtprelay=yes aleg_rtprelay_interface=wan rtprelay_interface=lan This works for voice calls, but not for T.38. The first thing notice is some values get changed in the SDPs involved in the re-INVITE transaction. Specifically, it adds a=sendrecv and a=direction:both to the INVITE's SDP. The 200 OK's SDP already has a=direction:passive, but Sems changes this to a=direction:both. I don't think either is important to the issue but I mention them for completeness. Both gateways agree on the T.38 session. At the beginning of the new media flow I see this error: [parse, AmRtpPacket.cpp:112] ERROR: bad rtp packet (hdr-size=12;pkt-size=8) ! As best I can tell there isn't anything wrong with the RTP / UDPTL flow to cause this. I see traffic continuing upstream on the wan-side, specifically the DIS portion of the v.21 traffic repeating several times. The gateway on the far-end does not respond so it seems like there is something wrong with some aspect of the traffic. These two fax machines and gateways communicate well using T.38 without Sems in the middle. How can I debug this? Regards, Jeff
_______________________________________________ Sems mailing list [email protected] http://lists.iptel.org/mailman/listinfo/sems
