On Fri, 23 Nov 2007, Clive D. W. Feather wrote:

In article
, Mizter T
writes

May I enquire what the forthcoming central portability database is all
about?

Let's suppose you started with a BT line, but then moved to Virgin Media
while keeping your number. At present, when somebody calls you, the call
is sent to the BT exchange handling your old (now removed) line. This
notes that you're a ported customer, sticks a prefix (say 527724) on the
front of your number, and re-injects the call into the trunk network.
This prefix means that it's now routed to the VM exchange handling your
line, which can deliver the call to you. This technique is called
"onward routeing" and is relatively inefficient.

Hang on, what is it that gets re-routed? Do the ATM cells carrying the
actual audio signal go to the old exchange and get forwarded, or is it
some kind of initial handshake that gets forwarded, with the audio then
travelling over a sensible path? I can see that the former would indeed be
very inefficient, but the latter doesn't seem to bad.

If it is the former, is this central database the right solution? How
about a redirection, like the HTTP 3xx status codes, where the old
exchange responds to the call setup handshake by saying "sorry old bean,
that number's now at exchange 527724", and the initiator then gets in
touch with the right exchange directly? I suppose that would still mean
the old exchange gets bothered by lots of annoying requests, which the
centralised approach avoids, but i imagine the aggregate cost would be
similar to that of using the database - less, in fact, since the majority
of calls will avoid the redirection step, whereas with the database, every
call makes a round-trip to it.

Although the database approach is more general and elegant, and probably
makes more sense long-term. It's basically adopting the internet's model
of having domain names sit on top of IP numbers with the DNS in the
middle. As the old computer scientists' proverb goes, there's no problem
that can't be solved with one more layer of abstraction.

tom

--
The whole of science is nothing more than a refinement of everyday
thinking. -- Albert Einstein