W3C home > Mailing lists > Public > ietf-http-wg@w3.org > April to June 2007

on the subject of DHCPv4

From: David W. Hankins <David_Hankins@isc.org>
Date: Tue, 26 Jun 2007 12:56:31 -0700
To: ietf-http-wg@w3.org
Message-ID: <20070626195631.GC3771@isc.org>
Sorry that I was on vacation during the thread on protocol support
for intercepting proxies;

  http://lists.w3.org/Archives/Public/ietf-http-wg/2007AprJun/0365.html

I don't really have a lot to add, but there were some DHCP bits I
wanted to respond to, so people have a clear(er) picture.


Adrian Chadd:
> Me, I'd prefer to see the proxy discovery draft properly worked into
> an RFC as there are -plenty- of instances of WPAD being used in the
> real world these days. It'd also be nice to have it extensible to proxy
> other protocols, such as P2P client proxy discovery (when P2P caches
> become all the rage, that is..)

I'd support this move with whatever limited time I can donate.  Even
just to document 252's use properly.

There was a fellow not long ago suggesting a DHCP option to carry
arbitrary basic proxy information.  I don't know what happened to him,
he seems to have just disappeared.

The idea wasn't a bad one, but I wondered if a "standard DHCP option"
for this practice would need to fill WPAD's considerably bulky shoes
to have a chance to be adopted by operators?


Henrik Nordstrom:
> And is why DHCP allows for larger messages (up to 64KB), if the client
> says it supports it..

Let's not enter into false advertising...

Trying not to bore you all with the details, or the matrix of
variables involved, the realistic answer is that today you can
probably use DHCP packets that are, at most, the same size as your
network's MTU (no fragments).  You might run into problems with very
old relay agents, but it's pretty unlikely today.

This is another problem area of DHCP protocol work that is easily
solved, but doesn't seem to be very pressing...option overloading
is very rare indeed (I've only ever heard of Sun network bootstraps
use it, as they can tend to include lots of big vendor options), so
no one is approaching the byte limit yet.


There were some other messages in there about DHCPv4's future.  I
don't mean any offense by not including and responding point by
point, it would just take too many paragraphs to try and respond to
all of it.

It took us 10 years to consume 128 options - and most of those were
the early frivolous 'gold rush' of options that sounded cool but
turned out not to be frequently used.

Suffice to say, I'm not the least bit worried, and I do DHCP protocol
work for a living.

Even if IPv6 is never adopted, there are still things we can do to
extend DHCPv4 well beyond 255 options, or even 64KB.

The sky is not falling.

-- 
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

Received on Tuesday, 26 June 2007 19:56:39 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 06:50:10 GMT