Summary:ASTERISK-03878: Incorrect handling of: 183 Session in progress. No early media.
Reporter:alladin (alladin)Labels:
Date Opened:2005-04-06 11:56:03Date Closed:2011-06-07 14:10:49
Versions:Frequency of
Environment:Attachments:( 0) debug.txt
Description:I think there's something wrong on how 183 is been handled. If I'll make a call from a sip client to asterisk and I forward that call to a sip gateway (via Dial()), it's impossible to get early media bridged between the client and the gateway (unless Answer() has been put in the dialplan, but that fsck up the CDR records :)
Comments:By: Kevin P. Fleming (kpfleming) 2005-04-06 14:05:20

As the bug guidelines request, please provide examples of your configuration, along with a complete 'sip debug/set verbose 255/set debug 255' trace.

By: alladin (alladin) 2005-04-07 04:25:53

I've attached a text file containing full debug.

By: Michael Jerris (mikej) 2005-04-07 12:45:06

We need configuration details as well please

By: Mark Spencer (markster) 2005-04-11 00:31:50

We need feedback here.  Also please confirm if this is an issue in CVS head, and please attach an ethereal dump.  Thanks.

By: Olle Johansson (oej) 2005-04-12 10:26:16

Reading the code "it should work" :-) We need serious feedback here to have something to dig into.

By: alladin (alladin) 2005-04-12 10:50:37

As I've entered, the bug is present in 1.0.7-stable. Haven't tried CVS-HEAD. This is clearly written in the table at the top of this page:

CVS HEAD/Stable or Release?   Release
CVS Date or Release Version 1.0.7

BTW, due the lack of time I can't provide an ethereal dump. Are there other quick tests I can do?

By: alladin (alladin) 2005-04-12 10:53:03

Regarding configuration, in the dialplan I've something like:

exten => _800.,1,SetCallerID(xxxxxx)
exten => _800.,2,Dial(SIP/${EXTEN}@xxxxxx)
exten => _800.,3,Congestion

By: Clod Patry (junky) 2005-04-12 11:55:11

sure, what you can if you wanna help is test with HEAD.
all patches here are for HEAD, not for stable directly.

By: Kevin P. Fleming (kpfleming) 2005-04-13 00:57:13

junky: that is not true, patches that correct bugs in stable are more than welcome here.

By: Kevin P. Fleming (kpfleming) 2005-04-29 13:10:09

The dialplan snippet you posted is _not_ what was used to handle these calls. From your log:

Apr  7 11:16:25 DEBUG[10092]: Launching 'SetVar'
Apr  7 11:16:25 VERBOSE[10092]:     -- Executing SetVar("SIP/6554000000-b52d", "LIMIT_WARNING_FILE=beep") in new stack
Apr  7 11:16:25 DEBUG[10092]: Launching 'SetCallerID'
Apr  7 11:16:25 VERBOSE[10092]:     -- Executing SetCallerID("SIP/6554000000-b52d", "5301392") in new stack
Apr  7 11:16:25 DEBUG[10092]: Launching 'Ringing'
Apr  7 11:16:25 VERBOSE[10092]:     -- Executing Ringing("SIP/6554000000-b52d", "") in new stack
Apr  7 11:16:25 VERBOSE[10092]: Transmitting (NAT):
SIP/2.0 180 Ringing
Via: SIP/2.0/UDP;branch=z9hG4bKa7b4.87d4d9a7.0;received=;rport=5060
Via: SIP/2.0/UDP;received=;branch=z9hG4bK_0.a8a21029
From: <sip:6554000000@sip.cheapnet.it>;tag=1112865392.734
To: <sip:0585255362@sip.cheapnet.it;user=phone>;tag=as335ade2f
Call-ID: 90977af5
CSeq: 14259 INVITE
User-Agent: Asterisk PBX
Contact: <sip:0585255362@>
Content-Length: 0

Your dialplan is executing the Ringing() app before using Dial() to send a call to your provider. That forces Asterisk to start early-media to the phone, and send audio ringing. Anything else that comes from the provider will be ignored (or at best, mixed in with Asterisk's ringing tones).

If you can replicate this problem using a dialplan that does _not_ generate any early media or indications itself, please reopen the bug.

By: alladin (alladin) 2005-05-03 02:56:32

I can assure you that the problem exists even if I take ring() outside the loop. That was probably a test (I can't really remember) to see if I was been able to resolve the issue myself. Obviously, the trick hasn't worked.

By: Michael Jerris (mikej) 2005-05-03 07:35:47

Ok, lets try this again.  We need the following: examples of your configuration, along with a matching complete 'sip debug/set verbose 255/set debug 255' trace and testing if this behavior *also* exists on current head.

By: twisted (twisted) 2005-05-25 16:44:41

alladin - is this still an issue?  If so, please provide the informaiton requested, or we will not be able to help you.

By: Michael Jerris (mikej) 2005-05-25 19:09:10

alladin-  Please reopen this when you have the requested information.