On vrijdag, mei 9, 2003, at 01:08 Europe/Amsterdam, Tony Li wrote:
I would prefer that we avoided having a stateful mapping mechanism.
It all flows from having more than one address with different
reachability properties. At some point, someone has to decide which
address to use for a particular purpose, and it had better be the
"right" address most of the time or we're worse off than being
single-homed.
It's unnecessary and certainly more complication than we need.
It doesn't have to be complicated: something as simple as publishing
all the addresses in the DNS and/or in an option in the first packet
along with a simple algorithm that figures out which address seems to
be working best is all we need. (But it may not be all we want.)
What kind of stateless mechanism do you have in mind? Obviously a
stateless solution would be preferable over a stateful one.
What worries me as ex-operator is that state in a network comes at a
cost. Normally state is actually among the most costly components of
network design. So, I agree that I would like to avoid any stateful
mapping. Now, that said I agree that a stateless mechanism is going to
harder to design and implement.