Hi Vishai, I made some commercial IP phone, and did a lot of interoperation tests with CSCFs in view of a UE In those situations, always transition between UDP and TCP was an issue. Because depends on a services, SIP message could have various size over MTU. But if an entity sends the sip message which size is over MTU without changing transport from UDP to TCP between a UE and CSCF, it will be fragmented and the message won't be serviced by UE or CSCF. This is a real-case and it's inevitable within real-services. I know that there are some ambiguous aspects in 3261 about that. Lastly there is an effort about this, please refer to this draft-gurbani-sip-large-udp-response-00.txt I'm not sure this one is the latest :)
Regards, ---------------------------------------------------------------------- Young-Joon Ryu (Alex) Team Manager IMS Platform Team Nable Communications, Inc. Tel : +82-2-3288-4334 (Ext:236) Mobile: +82-10-7701-8853 Fax: +82-2-558-8325 Email : [EMAIL PROTECTED] http://www.nablecomm.com ----------------------------------------------------------- -----Original Message----- From: Vishal Mathur [mailto:[EMAIL PROTECTED] Sent: Thursday, November 15, 2007 7:07 PM To: 류 영준; [EMAIL PROTECTED]; [email protected] Subject: RE: [Sip-implementors] How to determine transport to use from Routeheader Hi Alex, I am not sure that approach of send sip messages our TCP is correct if the original transport was UDP. This is because sender might not be having any connection open on TCP at all. That means sender has only connection open for UDP transport. I think SIP RFC allow use to implement SIP stack only on UDP or TCP or even both. So if the implementation is only with UDP the above approach will fail. Have any body tried sending reply to TCP port for Cisco ATA or IP phone when original message was generated on UDP? Does it work? Regards, Vishal Mathur Jabber ID: [EMAIL PROTECTED] Symantec Corporation www.symantec.com _________________________________ Office: +91-20-66067655 -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of ? ?? Sent: Thursday, November 15, 2007 2:15 PM To: [EMAIL PROTECTED]; [email protected] Subject: Re: [Sip-implementors] How to determine transport to use from Routeheader Hi all. In addition to below, if the total size of sip message is over a MTU in constructing, you have to change the transport to TCP even though the original transport was UDP. -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of 라스토기 Sent: Thursday, November 15, 2007 10:33 AM To: [email protected] Subject: Re: [Sip-implementors] How to determine transport to use from Route header Hi, if route has sip then it is UDP, if sips then TLS and sip with transport as TCp then TCP. I hope it helps... Vipul Rastogi Engineer, Business Management Team Telecommunication Network Business Samsung Electronics CO, LTD Suwon P.O.BOX 105, 416 Korea 442-600 MO 010-9530-0354 Hi, Can anyone please tell me How to find out the transport protocol to send the request, when we have to use the Route header for determining next hop. Suppose we have a Route header with the value Route: <sip:10.58.2.21:5060;lr>, <sip:10.48.2.40:5061;lr> Now because the first route value contains the lr parameter so the IP and Port of the next hop will be 10.58.2.21 and 5060 respectively, But How to find out the transport to be used in this case. Regards, Jitendra. _______________________________________________ Sip-implementors mailing list [email protected] https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors _______________________________________________ Sip-implementors mailing list [email protected] https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors _______________________________________________ Sip-implementors mailing list [email protected] https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
