[asterisk-users] PRI TBCT - Practical Experience, Anybody?

[Tilghman Lesher]

On Friday 15 August 2008 22:16:34 Jay R. Ashworth wrote:
> On Fri, Aug 15, 2008 at 02:49:17PM -0500, Tilghman Lesher wrote:
> > > To be more clear, what I'm after is to have *someone else besides me*
> > > place calls out their PRI, and then TBCT those placed calls to my DN.
> > >
> > > By the time the calls get to me, they should just be standard phone
> > > calls.
> > >
> > > So I expect the call-placing-party to need TBCT, but not me.
> > >
> > > > I believe that only DMS100, NI2 and 5ESS PRI signalling protocals are
> > > > capable of TBCT with the current zaptel code-base. Also, the two B
> > > > channels involved in the TBCT have to use the same D channel.
> > >
> > > And I'm probably not concerned with whether Asterisk can deal with
> > > TBCT, because Asterisk probably won't be involved at that stage; just
> > > once the call's transferred to me.
> > >
> > > But before I inquire of said second party whether they *can* do that, I
> > > wanted to confirm it was possible.
> >
> > 2BCT works when the telco originates the call and Asterisk is hairpinning
> > the call back out the same PRI circuit.  However, Asterisk does not
> > support the opposite direction.  That is, a call originated from Asterisk
> > that comes back in via the same PRI circuit cannot be 2BCT.  I'm not
> > certain whether this is a limitation of Asterisk alone or of the
> > protocol, but it cannot be done.
>
> I'm not sure we're not talking at cross purposes, here, Tilghman...
>
> but TBCT is an instruction to an end-office that sent you a call to
> yank it back off your timeslot and forward it along to someone else.
>
> There's no hairpin involved: the point of TBCT is that you tie up *0*
> timeslots instead of 2, to forward a call.

There is a hairpin involved.  The call (for several milliseconds at least) is
using two channels on the PRI before the 2BCT succeeds, and then the call
no longer takes up any channels.  It is only when the PRI detects the hairpin,
through the native bridge code that it is able to detect that the call is
eligible for 2BCT.

> Why would an Asterisk instance call itself on the same span?

Very simple.  Call your main number, and if you don't have special logic
in your internal dialplan context to handle that, the call will go out to the
telco and dutifully come right back in on the same circuit.

> > Similarly, Asterisk cannot complete a 2BCT request, if Asterisk is on the
> > NET side of the PRI circuit.  That might could be added in the future,
> > but it is not supported now.
> >
> > So in summary, Asterisk can request 2BCT, but it cannot perform a 2BCT if
> > requested from the other side.
>
> Nothing can perform a TBCT unless it's a PRI server, not a client; it's
> function of 5ESS's and DMSen; you have to be an SS7 speaker to do it in
> the first case.

I don't think that's the case.  Matt would know more, and his reply suggests
that it certainly would be possible for Asterisk to do this.  SS7 is not at
all required here.

-- 
Tilghman