Re: reuse callID by proxy

Thanx, I have checked out NIST stack source, call-id is used to dig
diglog out, so does it work? i mean the parameter -p.

And if there are more than one proxy on the dialog path, it could be
XXX -p -p -p ...... :)


urs Lebing


On Thu, 2006-09-28 at 04:35 -0700, Michael kr wrote:
> When you create a statefull proxy (or B2BUA - back to
> back UA) you basically create a new transaction for
> whatever you get from the proxy input and send it
> over. My experience is that using the call-id is the
> best way to map between the original and proxy
> transactions.
> 
> > The question is
> > that a stateful proxy
> > session based on NIST Stack has 1 or 2 dialog leg.
> The answer is 2 dialog legs.
> 
> > The other question is, a SIP Proxy is only on
> > transaction level or nor?
> > only map between CT and ST instead of dialog to
> > dialog. So any idea to
> > implement it based on NIST Stack.
> The map is between UA input and proxy transaction as
> CT and ST can be interchange between the parties.
> Again: make the map using the call-id so whoever comes
> from xxxx-p mapped to xxxx and vise versa.
> 
> Michael
> 
> 
> --- Lebing Xie <[email protected]> wrote:
> 
> > On Wed, 2006-09-27 at 05:57 -0700, Michael kr wrote:
> > > Just use the original call-id with some extension
> > > (like "-p") to identify the proxy transaction.
> > > 
> > 
> > Does NIST stack support it? I checked source, seemed
> > not(?) 
> > is there any spec about it, thanx. The question is
> > that a stateful proxy
> > session based on NIST Stack has 1 or 2 dialog leg.
> > 
> > The other question is, a SIP Proxy is only on
> > transaction level or nor?
> > only map between CT and ST instead of dialog to
> > dialog. So any idea to
> > implement it based on NIST Stack.
> > 
> > urs Lebing
> > 
>