On Thu, May 29, 2008 at 12:29 PM, M. Ranganathan <[EMAIL PROTECTED]> wrote:
> On Thu, May 29, 2008 at 12:15 PM, Robert Joly <[EMAIL PROTECTED]> wrote:
>>>
>>> I have the following situation that can result in a resource leak
>>>
>>> Call setup between music on hold ( or some long playing
>>> stream that does not hang up ) and caller who calls in via the ITSP.
>>> Remote end drops call without bye.
>>>
>>>
>>> I have to do some liveness testing and destroy the stream.
>>> The question is how to do this?
>>>
>>> Initially, I thought RTP would be good enough but then
>>> potentially the phone is not sending any RTP.  Can I rely
>>> upon the phone sending RTCP?
>>> I would like to use  RTCP receiver reports for liveness
>>> testing of the remote endpoint and hang up automatically  if
>>> I dont see any RTCP reports from one or more of the
>>> participants of the call.  Is this a sound strategy?
>>>
>>
>> Although mandatory, it is not uncommon to find SIP endpoints that do not
>> support RTCP.  I would therefore claim that relying on RYCP will not
>> yield the desired results.  Have you looked into adding session timers
>> support to your B2BUA (RFC4028).  Session timers can work even if the
>> other end does not support them.
>>
> Thanks. I guess the safest option would be to use reINVITE ( with SDP
> ) as a in Dialog target refresh because not all ITSPs would
> necessarily support OPTIONS.

UPDATE is another possibility I suppose but that pre-supposes support
for this method.

Ranga

>
> Ranga.
>
>
> --
> M. Ranganathan
>



-- 
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