when sems receives voice call invite from ua behind nat that has
direction:active in its sdp, like this:
Session Description Protocol
Session Description Protocol Version (v): 0
Owner/Creator, Session Id (o): foo.bar 0 0 IN IP4 192.168.1.
68
Session Name (s): -
Connection Information (c): IN IP4 192.168.1.68
Time Description, active time (t): 0 0
Media Description, name and address (m): audio 8500 RTP/AVP 18 4 112 100 101 0
8 106
Media Attribute (a): sendrecv
Media Attribute (a): rtpmap:18 G729/8000
Media Attribute (a): rtpmap:4 G723/8000
Media Attribute (a): rtpmap:112 AMR/8000
Media Attribute (a): rtpmap:100 EG711U/8000
Media Attribute (a): rtpmap:101 EG711A/8000
Media Attribute (a): rtpmap:0 PCMU/8000
Media Attribute (a): rtpmap:8 PCMA/8000
Media Attribute (a): rtpmap:106 telephone-event/8000
Media Attribute (a): direction:active
sems includes direction:passive in 200 OK and starts to correctly send
rtp packets to the ip address/port of received rtp packets.
however, then sems receives a voice+video call invite like this:
Session Description Protocol
Session Description Protocol Version (v): 0
Owner/Creator, Session Id (o): foo.bar 0 0 IN IP4 192.168.1.68
Session Name (s): -
Connection Information (c): IN IP4 192.168.1.68
Time Description, active time (t): 0 0
Media Description, name and address (m): audio 8500 RTP/AVP 18 4 112 100 101 0
8 106
Media Attribute (a): sendrecv
Media Attribute (a): rtpmap:18 G729/8000
Media Attribute (a): rtpmap:4 G723/8000
Media Attribute (a): rtpmap:112 AMR/8000
Media Attribute (a): rtpmap:100 EG711U/8000
Media Attribute (a): rtpmap:101 EG711A/8000
Media Attribute (a): rtpmap:0 PCMU/8000
Media Attribute (a): rtpmap:8 PCMA/8000
Media Attribute (a): rtpmap:106 telephone-event/8000
Media Description, name and address (m): video 8600 RTP/AVP 34
Bandwidth Information (b): AS:300
Media Attribute (a): sendrecv
Media Attribute (a): rtpmap:34 H263/90000
Media Attribute (a): direction:active
it does not include direction:passive in 200 OK and keeps on sending rtp
packets to 192.168.1.68 even after receiving first voice rtp packet from
public ip of the ua.
is this a bug in sems or is it a bug in sr nathelper_k
fix_nated_sdp("1") implementation, i.e., does that function place
direction:active line to only one place, when it should place it in two
places (in both voice and video media descriptions)?
-- juha