On Wed, Nov 01, 2006 at 11:21:18AM -0700, Ted Lemon wrote:
> I'd just like to point out that by introducing this
functionality in a
> failover environment, you are roughly doubling the load
on each DHCP
> server. So while I'm sympathetic to your argument that
the relay agent
> shouldn't be maintaining state, there is a substantial
cost involved.
I think it is important to enter into state-keeping on the
relay
where it is expressly necessary to meet an ends.
I'm growing more aware in the absence of alternative
solutions I had
dreamed for this, there is going to need to be some state
keeping on
relays to continue to afford failover implementations a
'load
balancing' incentive.
I also think it's important to make sure that if state
keeping is
mandated for one purpose, it is not mandated for the general
purpose.
That is, server-override should require no state keeping by
the relay.
Failover support on a relay implementing server override
might require
state keeping, no matter how much I wish it wouldn't.
--
David W. Hankins "If you don't do it right the first
time,
Software Engineer you'll just have to do it again."
Internet Systems Consortium, Inc. -- Jack T. Hankins
_______________________________________________
dhcwg mailing list
dhcwg ietf.org
https://
www1.ietf.org/mailman/listinfo/dhcwg
|