[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: PI/metro/geo [Re: The state of IPv6 multihoming development]

On Friday, Nov 1, 2002, at 13:08 America/Montreal, Tony Hain wrote:
It is the 'cost optimized' part that makes the difference. Which pockets
are we talking about protecting? I am not saying we should favor any
particular pocket, just asking if we are doing it unintentionally by the
representatives around the table.
All the pockets.  Trying to shift costs onto ISPs just means they
are forced to raise rates and shift them back to customers.  Terribly
sorry, but that's how the economics work in this case.  And the
part I strongly dislike is that those raised rates propogate much more
widely than the subset of sites that cause the ISP costs -- and harm
the general growth and health of the global Internet.


This model was probably possible 4 years ago, but at this point I doubt
tweaking the pseudoheader can even be on the table.
If router vendors had a knob that enabled separation of location/identity
and the IDR infrastructure was compatible with that, I bet a whole lot
of enterprise users would enable that knob eagerly in order to obtain
multi-homing and backup and other things they desire.

It might be interesting to run that experiment, though obviously the
implementation specifics would need to be worked out first.

Also, 'freely
replaced' usually sets off the alarm bells of the spoof-sensitive
security types.

I've been accused of Internet paranoia in the past.  Please note that
I disagree with the assertion that such paranoia is reasonable in the
case of separating location from identity.  Unreasonable paranoia
exists no matter what one does, so ought not be a variable in the
decision process.
That does not mean we GSE is hopeless, just that we will
probably have to use something like MIPv6 to mask the label swapping.
In the abstract, there is not much difference between GSE and a Care-of
Address. So for routing & packet forwarding, if the CoA had finer
grained pattern replacement happening the end systems would not be aware
of it. The missing link would be letting the host know what the current
publicly visible address is so it could pass that to the CN.
I'm happy to discuss whichever implementation details. I just think that
from an architectural perspective, we need to separate location and identity
to make any significant headway here.

Ran