[VoiceOps] Strange Call Flow

[Scott Berkman]

Yes its "legal", a RFC interpretation but not very popular.   It is usually
referred to as "Multiple 18X response" and it can normally be turned off.
What they are doing here is starting w/o early media (calling side provides
ringback) and then trying to change over to early media without a re-invite,
possibly to provide the ring-back or a treatment.

Here are some links:
http://tools.ietf.org/html/draft-sen-sip-earlymedia-00
http://www.ietf.org/rfc/rfc3261.txt

Technically I think you can keep sending provisional responses (1XX) as many
times as you like until a final response (2XX, 3XX, etc) is sent.  One
example of course is 100 Trying followed by a single 18X.  the 100 trying is
optional.

	-Scott

-----Original Message-----
From: voiceops-bounces at voiceops.org [mailto:voiceops-bounces at voiceops.org]
On Behalf Of Tim Donahue
Sent: Friday, October 15, 2010 1:49 PM
To: voiceops at voiceops.org
Subject: [VoiceOps] Strange Call Flow

I'm seeing a strange call flow from a client with a Cisco CUBE setup.
This is happening when they attempt to forward a call from their system to a
PSTN line.

SBC                          CUBE
 | ----------INVITE---------> |
 | <------100 Trying--------- |
 | <------180 Trying--------- |
 | <--183 Session Progress--- |
 | <--------200 OK----------- |

I've never seen a call setup that switched from a 180 Ringing to a 183
Session Progress mid-setup before, and it seems to be doing funny things to
the media transport through our SBC (we see audio come in from one of the
call legs but it doesn't get forwarded to the other call leg).  Is this a
legal call flow that we should add to our tests when evaluating new
hardware?

Has anyone seen problems like this before?

Tim
_______________________________________________
VoiceOps mailing list
VoiceOps at voiceops.org
https://puck.nether.net/mailman/listinfo/voiceops