[Serusers] Re: [Serdev] Problem with Via Branch/Strange Looping Problem with ACK's

Zeus Ng zeus.ng at isquare.com.au
Wed May 4 17:27:41 CEST 2005


Alex,

I agree with you that "branch=0" is a bad implementation. A similar
discussion was raised last year in the sip-implementors mailing list.

http://lists.cs.columbia.edu/pipermail/sip-implementors/2004-September/00724
3.html

It illustrated the problem with examples. Maybe someone should look into
this issue.

Zeus

> -----Original Message-----
> From: Alex Mack [mailto:amack at fhm.edu] 
> Sent: Wednesday, 4 May 2005 10:32 PM
> To: Zeus Ng
> Cc: serdev at lists.iptel.org; serusers at lists.iptel.org
> Subject: Re: [Serusers] Re: [Serdev] Problem with Via 
> Branch/Strange Looping Problem with ACK's
> 
> 
> Hi Zeus!
> 
> Maybe there's a misunderstandiung: I for myself have no problem with 
> looping packets but I've problems with ACKs and Via-HF with 
> branch=0 - 
> and that's the UAS (SER) to be blamed for. The UA sends a 
> normal - and 
> in my eyes correct - ACK after an 200 OK, and this ACK is 
> forwarded by 
> SER with a Via-HF containing "branch=0".
> 
> So far I didn't get a clue how I could take influence on SER 
> writing bad 
> branch parameters - even if I wanted SER to do that. IMHO 
> branch=0 isn't 
> legal according to the metioned sections of RFC3261.
> 
> I've attached a ethereal plaintext export of a call. Packet 
> #13 is the 
> ACK from the initiating UA to SER, packet #14 is the bad ACK with 
> branch=0. Sorry for the crumbled view, but I have almost no 
> influence on 
> ethereal's export format.
> 
> Alex Mack
> 
> Zeus Ng schrieb:
> 
> >When I mention section 12.2, I was referring your UA, not SER. I do 
> >have a typo on alias as well.
> >
> >To better troubleshoot your problem, ser.cfg and debug would 
> help. By 
> >looking at your ngrep, I believe there is something wrong with your 
> >ser.cfg. The ACK from SER to Asterisk should not have a Record-Route 
> >header added.
> >
> >
> >
> >  
> >
> >>-----Original Message-----
> >>From: serdev-bounces at lists.iptel.org
> >>[mailto:serdev-bounces at lists.iptel.org] On Behalf Of Alex Mack
> >>Sent: Wednesday, 4 May 2005 8:29 PM
> >>To: Zeus Ng
> >>Cc: serdev at lists.iptel.org; serusers at lists.iptel.org
> >>Subject: Re: [Serusers] Re: [Serdev] Problem with Via 
> >>Branch/Strange Looping Problem with ACK's
> >>
> >>
> >>Hi Zeus!
> >>
> >>Zeus Ng schrieb:
> >>
> >>    
> >>
> >>>First, the ACK should follow section 12.2 (indeed 12.2.1.1),
> >>>      
> >>>
> >>Requests
> >>    
> >>
> >>>within a Dialog, not section 12.1.2 according to section
> >>>      
> >>>
> >>13.2.2.4, 2xx
> >>    
> >>
> >>>Responses.
> >>>------
> >>>The UAC core MUST generate an ACK request for each 2xx 
> received from 
> >>>the transaction layer. The header fields of the ACK are 
> constructed 
> >>>in the same way as for any request sent within a dialog
> >>>------
> >>> 
> >>>
> >>>      
> >>>
> >>I disagee: the SER is an UAS not an UAC.
> >>IMHO, the concernig chapters are 16.6 subpoint 8 ("16.6 Request
> >>Forwarding"; "8. Add a Via header field value") and 8.1.1.7 
> ("8.1.1.7 
> >>Via"). Both chapters state the UAS (SER) MUST add a Via 
> with a branch 
> >>parameter starting with the magic cookie "z9hG4bK".
> >>
> >>The ACK generated by the UAC looks good to me, anyway.
> >>
> >>    
> >>
> >>>Second, do you happened to have a line like this in your 
> SER config.
> >>>
> >>>aliase=your_proxy_name:5080 or aliase=your_proxy_ip:5080?
> >>>
> >>> 
> >>>
> >>>      
> >>>
> >>Are you sure about the trailing 'e' in "aliase"? I've two
> >>alias-definitions (one for the IP, one for the realm) in my 
> >>ser.cfg. But 
> >>they don't have the port added. I've already added the ports in the 
> >>listen statements.
> >>
> >>Alex Mack
> >>
> >>_______________________________________________
> >>Serdev mailing list
> >>serdev at lists.iptel.org http://lists.iptel.org/mailman/listinfo/serdev
> >>
> >>
> >>    
> >>
> >
> >
> >
> >  
> >
> 
> 




More information about the sr-users mailing list