This is quite a realistic situation e.g. 'BYE Crossing on the wire'. I personally prefer 1. Side A has issued a BYE and so should wait for the response before hanging up. But as soon as it receives a BYE, it should issue a 200 OK and terminate the call. But beware, all SIP transactions MUST terminate independently, so the UA should be willing to receive the 200 OK (silently absorb) for it's BYE.
Chris. >>>>-----Original Message----- >>>>From: Daniel Bar-Yakov [mailto:[EMAIL PROTECTED] >>>>Sent: 13 March 2003 09:42 >>>>To: [EMAIL PROTECTED] >>>>Subject: [Sip-implementors] BYE Glare Cleanup Question >>>> >>>>Greetings, >>>> >>>>I believe this question is a little different from the questions already >>>>raised regarding BYE glare situations. (Though being new to the list, I >>>>may be missing something.) I did find references to what kind of >>>>response >>>>to send in step (3) below. >>>> >>>>However, I'm looking for a recommendation as to what to do in step (4). >>>> >>>>Sequenced scenario: >>>> Side A sends a BYE request >>>> Side A receives a BYE request (rather than the expected BYE response) >>>> Side A duly sends a BYE response >>>> Implementation question point - >>>> >>>>Side A options for Step (4): >>>>Option 1: Side A clears the call, no longer waiting for the BYE response >>>>to >>>>its BYE request, >>>> OR, >>>>Option 2: Side A waits (or times out) for the BYE response to its >>>>request, >>>>before clearing the call. >>>> >>>>Option (1) acts on the established fact that both sides are interested >>>>in >>>>ending the call, so no more messaging is necessary. >>>>Option (2) lends better symmetry to both sides of the glare, since Side >>>>B >>>>could have experienced the exact same sequence. It is therefore likely >>>>that a BYE response is traveling somewhere between the application >>>>layers >>>>of Side A and Side B. >>>>I believe both options are compliant with the SIP spec. >>>> >>>>Thank you, >>>>Daniel Bar-Yakov >>>>NMS Communications - Israel >>>>www.nmscommunications.com >>>> >>>> >>>>_______________________________________________ >>>>Sip-implementors mailing list >>>>[EMAIL PROTECTED] >>>>http://lists.cs.columbia.edu/mailman/listinfo/sip-implementors _______________________________________________ Sip-implementors mailing list [EMAIL PROTECTED] http://lists.cs.columbia.edu/mailman/listinfo/sip-implementors
