Dr. Rodney G. McDuff
2006-Mar-06 19:15 UTC
[Asterisk-Users] Confusion about construction of RURIs from contact headers for BYEs generated by *
I'm a bit confused about how * constructs the RURI when it generates a BYE. For the situation where * send the initial INVITE it constructs the RURI for the BYE from the contact header of the 200 OK response which is well and good. However when * receives the initial INVITE it does not use the contact header contained within to construct the BYE's RURI but constructs it from scratch. This is of particular concern when one is using a SIP clients that only do TCP like office communicator (and of course a proxy in the middle to do UDP/TCP protocol conversion) . In these cases well-behaving UA puts a transport=tcp tag into the contact header of either the INVITE or the 200 OK. In the first scenario this tag is preserved and the BYE's RURI has the transport tag (which the proxy can use for protocol conversion). The second scenario loses the transport tag in the BYE's RURI (and the proxy has no idea that protocol conversion is needed) Is my understanding of the code correct and if so was this something that was just missed or a design decision. -- Dr. Rodney G. McDuff |Ex ignorantia ad sapientiam Manager, Strategic Technologies Group| Ex luce ad tenebras Information Technology Services | The University of Queensland | EMAIL: mcduff@its.uq.edu.au | TELEPHONE: +61 7 3365 8220 |
Olle E Johansson
2006-Mar-07 00:35 UTC
[Asterisk-Users] Re: [asterisk-dev] Confusion about construction of RURIs from contact headers for BYEs generated by *
7 mar 2006 kl. 03.15 skrev Dr. Rodney G. McDuff:> I'm a bit confused about how * constructs the RURI when it generates a > BYE. For the situation where * send the initial INVITE it > constructs the > RURI for the BYE from the contact header of the 200 OK response > which is > well and good. However when * receives the initial INVITE it does not > use the contact header contained within to construct the BYE's RURI > but > constructs it from scratch. This is of particular concern when one is > using a SIP clients that only do TCP like office communicator (and of > course a proxy in the middle to do UDP/TCP protocol conversion) . In > these cases well-behaving UA puts a transport=tcp tag into the contact > header of either the INVITE or the 200 OK. In the first scenario this > tag is preserved and the BYE's RURI has the transport tag (which the > proxy can use for protocol conversion). The second scenario loses the > transport tag in the BYE's RURI (and the proxy has no idea that > protocol > conversion is needed) > > Is my understanding of the code correct and if so was this something > that was just missed or a design decision. >This seems to be a bug we have to fix. Some advice: - Please do not crosspost to two mailing lists - Always mention version of Asterisk you are using when reporting a problem - Open a bug report for this in the bug tracker, bugs.digium.com Thanks /Olle
Olle E Johansson
2006-Mar-07 00:46 UTC
[Asterisk-Users] Re: [asterisk-dev] Confusion about construction of RURIs from contactheaders for BYEs generated by *
7 mar 2006 kl. 03.15 skrev Dr. Rodney G. McDuff:> I'm a bit confused about how * constructs the RURI when it generates a > BYE. For the situation where * send the initial INVITE it > constructs the > RURI for the BYE from the contact header of the 200 OK response > which is > well and good. However when * receives the initial INVITE it does not > use the contact header contained within to construct the BYE's RURI > but > constructs it from scratch. This is of particular concern when one is > using a SIP clients that only do TCP like office communicator (and of > course a proxy in the middle to do UDP/TCP protocol conversion) . In > these cases well-behaving UA puts a transport=tcp tag into the contact > header of either the INVITE or the 200 OK. In the first scenario this > tag is preserved and the BYE's RURI has the transport tag (which the > proxy can use for protocol conversion). The second scenario loses the > transport tag in the BYE's RURI (and the proxy has no idea that > protocol > conversion is needed) > > Is my understanding of the code correct and if so was this something > that was just missed or a design decision. >This seems to be a bug we have to fix. Some advice: - Please do not crosspost to two mailing lists - Always mention version of Asterisk you are using when reporting a problem - Open a bug report for this in the bug tracker, bugs.digium.com Thanks /Olle _______________________________________________ --Bandwidth and Colocation provided by Easynews.com -- asterisk-dev mailing list To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-dev