Hi Shaun,
I have noticed the same thing. "media_timout" in conjunction with WebRTC does not work at all.
If someone has a solution, that would be really fine.
Currently I have helped myself by re-enabling the following in the code:
diff --git a/src/switch_rtp.c b/src/switch_rtp.c
index 40d8978..aada64a 100644
--- a/src/switch_rtp.c
+++ b/src/switch_rtp.c
@@ -854,8 +854,8 @@ static switch_status_t ice_out(switch_rtp_t *rtp_session, switch_rtp_ice_t *ice)
if (elapsed > 30000) {
switch_log_printf(SWITCH_CHANNEL_SESSION_LOG(rtp_session->session), SWITCH_LOG_WARNING, "No %s rtp_session->last_stun = switch_micro_time_now();
- //status = SWITCH_STATUS_GENERR;
- //goto end;
+ status = SWITCH_STATUS_GENERR;
+ goto end;
}
}
@@ -897,7 +897,7 @@ static switch_status_t ice_out(switch_rtp_t *rtp_session, switch_rtp_ice_t *ice)
ice->sending = 3;
- // end:
+ end:
READ_DEC(rtp_session);
return status;
(END)
The code has the consequence that the session is cleared if no more media comes for 30 seconds.
With kind regards
Alex
I'm not sure if others have had similar experiences but for us the media_timeout variable does not work as expected.
-
media_timeout on a call that supports video but with-out video will fail.
-
media_timeout_audio works in some instances, in others the timeout period is ignored so the call will timeout almost immediately after RTP stops.
-
media_hold_timeout_audio doesn't seem to work at all, calls that are on hold never timeout.
Why is the SIP profile parameter 'rtp-timeout-sec' depreciated? It's much simpler to apply this per SIP profile than it is per call.