Re: (PM) hunt group rollover when all modems are busy (fwd)

Derric Scott (dtscott@scott.net)
Sun, 9 Nov 1997 01:19:01 -0600 (CST)

Michael's description is pretty much exactly what I determined here (in the
US) and we're off a 5ESS. Could it be a programmable parameter in the
5ESS? Apparently (from both the telco and from Livingston Tech sup) the
PM3 returns: "Release complete, user busy - cause code of 17" in the
given situation. The Livingston caller I talked to indicated that there
wasn't really a "standard" code to return for the desired behavior and
then told me that they use the Cause Code 17. It would then seem to be up
to the switch to be able to "do the right thing" upon receipt of the Cause
Code 17.

I guess if you are fully populated with modems and if there is no
re-ocurrance of the ADMIN sort of bugs, then it will work ok...

Derric

> In the USA, Nortel DMS switches have the same problem but the 5ESS does
> NOT. The 5ESS can rehunt if the NAS reports a requested facility not
> available message over the D channel back to the switch.
>
> Scooter
>
>
> At 04:14 PM 11/7/97 +0100, Miquel van Smoorenburg wrote:
> >In article <3.0.3.32.19971107153423.03930aa0@pop3.a1plus.at>,
> >Michael Haberler <mah@austria.eu.net> wrote:
> >>The PM3 sends a busy signal to the ISDN switch when all analog modems are
> >>connected, and the next analog call comes in. The CW is that the telco
> >>switch now should rollover the call to the next PRI.
> >>
> >>I just had a long conversation with our Austrian PTT folks here on how the
> >>switches they use (Nortel, Siemens) handle the sitution. The answer is
> short:
> >>
> >>There seems to be no obvious way (for them..) to redirect a call to the
> >>next PRI in a hunt group if a PRI isnt fully busy but all modems 'behind'
> >>it are.
> >>
> >>This is the problem (*they* say):
> >>
> >>If a call reaches *the telco switch*, the switch *first* negotiates a
> >>B-channel with the PM3 to allocate. If the PRI is full, fine - switch rolls
> >>over to next PRI in hunt group since allocation fails, no problem here.
> >>
> >>If the PRI is *not full*, the B-channel allocation succeeds. Only *after
> >>the allocation*, the call setup message is presented on the D-Channel, with
> >>the bearer capability (voice or 64k). And only now the Pm3 can tell wether
> >>she can handle it (depending on all modems busy or not). Our PTT folk says
> >>if the get a reject at this point that was it for the call - there's no way
> >>to rollover to the next PRI - they say after b-channel allocation the call
> >>is 'committed' to this PRI (in transactional speak). Not good.
> >>
> >
> >Aha - so that's the reason for that. I've talked to at least 10 different PTT
> >folks here in Holland but only about one third understood my question and
> >no-one came up with a satisfactory answer as to why they couldn't do it.
> >
> >Another approach is to see if it is possible to get 2 different telephone
> >numbers, on the same set of lines. One group would roll from top to bottom
> >(ISDN) and the other from bottom to top (modems). Then just fully load
> >the first few portmasters with modems.
> >
> >Ofcourse, the Dutch PTT also claims this is impossible without satisfactory
> >explanation (hopefully we'll have better luck with Enertel)...
> >
> >Mike.
> >--
> > Miquel van | Cistron Internet Services -- Alphen aan den Rijn.
> > Smoorenburg, | mailto:info@cistron.nl http://www.cistron.nl/
> >miquels@cistron.nl | PTT's Het Net: Surfen in de gootsteen!
> >-
> >To unsubscribe, email 'majordomo@livingston.com' with
> >'unsubscribe portmaster-users' in the body of the message.
> >
> >
> - -
> To unsubscribe, email 'majordomo@livingston.com' with
> 'unsubscribe portmaster-users' in the body of the message.
>
> ------------------------------
> 

-- 
Derric Scott          Scott Network Services, Inc.         P. O. Box 361353
derric@scott.net           (205)987-5889               Birmingham, AL 35236


-
To unsubscribe, email 'majordomo@livingston.com' with
'unsubscribe portmaster-users' in the body of the message.