[Serusers] Ser2.0 wrong behaviour of the sl_reply () function?

Jiri Kuthan jiri at iptel.org
Mon Apr 30 11:53:16 CEST 2007


At 11:48 30/04/2007, tzieleniewski wrote:
>> At 16:35 25/04/2007, tzieleniewski wrote:
>> >Hi,
>> >
>> >I am using SER rel 2_0_0 from CVS.
>> >I encountered the following situation.
>> >My scenario was the following:
>> >
>> >UA -> Application Serwer -> SER -> UA
>> >
>> >1. UA sent an INVITE message to the application server.
>> >2. Application Server changed the Request URI and forwarded statelessly the INVITE message to SER.
>> >3. Everything would be ok but SER replied to the Application server instead of the client. 
>> 
>> It can be that you meant to say something different, but that's standard 
>> SIP behqviour that you send a reply to the most immediate upstream
>> entity (AS in this case).
>> 
>> >Application server changed only the Request URI and nothing else.
>> >Shouldn't SER sent the response according to the VIA header value??
>> 
>> You mean to say that the AS didn't add its own Via in there?
>> 
>Yes, the aim was to apply some external to SER modifications on the SIP massage and then return it back to SER. It was suppose to be something like invocation of some external logic outside the SIP routing procedures. 

I would say it is a bit slippery slope to do this ...


>> That appear kind of odd to me, I would begin fixing there.
>> 
>> But if you think there is possibly some
>> value in it, there is I believe an option in SER to send back literally
>> to what is printed in Via (as opposed to reversely sending to previous
>> hop's transport address). It is normally turned off, becasue Via content
>> is frequently useless (e.g., due to presence of NATs).
>> 
>Would be thanks full if you did:) 

... but if you really want to try, set reply_to_via=1.

-jiri 


--
Jiri Kuthan            http://iptel.org/~jiri/




More information about the sr-users mailing list