In your previous mail you wrote:
> => today a standard box can't run at 1GE then I suppose your
My box do. And Apple's new computer is shipping with GE interfaces for
instance...
=> standard boxes with a standard PCI can't go at 1Gbits/s.
There is a difference between to put a giga Ethernet board in a box
and to go at 1 Gbits/s. I know this will be fixed (I'd like ASAP)
but the gap between standard boxes and ASIC is enough to be able
to use ASICs for hard things at standard box speed. And I believe
this will be true in the long term then the issue is hard things
can be done only by edge devices and not in the core, but we know
that (as I said this is the IntServ/DiffServ lesson).
If you look at figures from Garter Group, IDC, DellOro, you see a trend that
GE will enter the LAN market very fast...
=> where? I know GE for campus backbones but GE is not very common for
hosts (for three reasons today: fibers are expensive (will change when
copper GE will be common), boards are expensive (will change when GE will
spread enough) and PCI bus is too slow (will change when chipsets will
be improved but (this is a shame, I agree) this seems to take time)).
just like FE entered the market
fast when it come and people used the same cabling but upgraded from 10 Mb
to FE.
=> the same cabling hypothesis seems less likely for GE.
> => I had the wish I missed it. You can't reuse the payload
> length because
> it is *not* free. You propose to create a very different protocol with
> a different (likely larger) header...
Yes - with a chance to make a deterministic implementation of it in asic, np
etc...
=> (I have not the answer) Is a discussion about a new protocol in the scope
of this list?
> > => I think nothing good of this. It is not the job of a
> > router to look at
> > inside packets (ESP will save us!)
This trend is driven from the operators and the router vendors ingrate more
and more into its boxes... You need to look into the packet at every edge
(backbone/WAN edge, MAN edge, LAN edge/access edge etc.) This is also a
natural way to go when you go up in speeds.
=> I don't understand, I believed you should do simpler and simpler things
when you want/need to be faster and faster, not the opposite.
> => I believe the core routers are not able to classify the
> packets then
> my position is a bit more than "in favour".
I was not talking about core routers...
=> oc-768 isn't for core routers? I don't understand your arguments...
[EMAIL PROTECTED]
--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page: http://playground.sun.com/ipng
FTP archive: ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------
