> From: "Tony Li" <[EMAIL PROTECTED]>
> If we go down the path of map-&-encap, we're effectively deciding to run
> on top of a connection oriented infrastructure.
I don't think I agree with this. It is certainly true that we're almost
building a new packet-switching system on top of an existing one - and this
certainly creates some problems (which I'm currently struggling with offline,
along with a number of other people).
I saw "a new packet-switching system" because at the ITR you have to keep
track of all the ETRs which are available to get to the ultimate destination;
decide which one to use; make sure it's up, and switch to a different one if
it's not; etc, etc. All sounds very familiar, doesn't it? :-)
But I don't at all see how treating the entire Internet as a NBMA network
means we're treating it like a "connection oriented infrastructure", any more
than when we did almost exactly the same thing, when we started running IP
over the old ARPANet. All we're doing is treating a composite NBMA network
(an internetwork, to be technical) as if it were a classic single physical
NBMA, which the 'routers' (xTRs in this case) use to communicate directly
with each other, without much knowledge of what's going on inside that 'black
box' NBMA they are using for data carriage.
What's troubling me much more (wearing both engineering and architectural
hats) is that we are, in effect, building a second packet-switching system on
top of the first, but separate from it - and with only the dimmest of plans
(if any, in most cases) to make sure that in the long run we have a path _out_
of that kludgy, exponentially-complex mess, back to something a little more
integrated, clean, flexible, powerful and robust.
> MPLS ... allows us to have a hybrid network architecture, where
> we can support both connectionless (i.e. packet switching) and
> connection oriented (i.e. circuit switching) styles.
Ever since this was first explicitly pointed this out to me (by Yakov), I've
always felt it was one of the strong points of MPLS (although I think it was
still called the older name when I had that chat), and of couse of this class
of architecture in general.
(Indeed, a system I proposed some time before provided both connectionless -
which I termed datagram - and flow-oriented service, on top of a
flow-oriented substrate. I use the term 'flow-oriented' because to many
people 'connetion-oriented' also means in a lot of other stuff, like
reliability, which I don't mean to include.)
Noel
_______________________________________________
rrg mailing list
[email protected]
https://www.irtf.org/mailman/listinfo/rrg
