On Tue, May 19, 2009 at 4:15 PM, Dale Worley <[email protected]> wrote:
> On Tue, 2009-05-19 at 15:47 -0400, M. Ranganathan wrote:
>> Hello,
>>
>> Attached is a trace for a test which is configured to do the following:
>>
>> Request from INVITE is directed to a dial plan (extension 501) which
>> parallel forks to two extensions 201 and 202. Each of those extensions
>> has an unconditional call forwarding to the SAME ITSP number. JAIN-SIP
>> detects the loop and responds to one of the call legs with LOOP
>> DETECTED ( Frame 91 ).
>>
>> However, by this time the ITSP (PSTN) phone is ringing.
>>
>> Shortly thereafter, SIPX PROXY sends a CANCEL. This CANCELs the
>> original call. Questions:
>
> The INVITE to sipxbridge is frame 86 at 19:20:49.
>
> The CANCEL to that invite is frame 122 which is at 19:20:59, which is 10
> seconds later.  This is enough to suggest that the CANCEL is due to an
> event which is asynchronous to the INVITE.
>
> The CANCEL propagates from CANCEL frame 116 at 19:20:59, which cancels
> the INVITE at frame 57 at 19:20:49.
>
> Generally, the conditions that cause a CANCEL to be created are set up
> not by the INVITE that it cancels, but by the parent INVITE.  In this
> case, looking at the *second* branch value of frame 57 and scanning
> upward, we discover the parent INVITE is frame 23, INVITE sip:us...@...,
> 19:20:29, which has "Expires: 30".  And CANCEL frame 116/122 is at
> 19:20:59, which is 30 seconds later.
>
> So the immediate cause of the CANCEL is that the INVITE to user1 has a
> 30 second expiration, and it takes 20 seconds before the INVITE
> propagates to sipXbridge.  That latter appears to be because the call
> goes to two phones before getting routed to sipXbridge.

Ah! Makes perfect sense.  I changed the expiration time and it worked
as expected.

Thanks !

Ranga

>
> Dale
>
>
>



-- 
M. Ranganathan
_______________________________________________
sipx-dev mailing list
[email protected]
List Archive: http://list.sipfoundry.org/archive/sipx-dev
Unsubscribe: http://list.sipfoundry.org/mailman/listinfo/sipx-dev

Reply via email to