On the record - debunking technical fallacies

> Dean has weighed in on topics such as router
architecture and the
> ubiquitousness of packet-based-load-balancing in
backbone networks, and
> been thoroughly wrong.

I never said that PPLB is ubiquitous (widely
used--for those not so used
to big words). I said that it is possible to see
it. And that if you see
it, it will not work with anycast TCP DNS.

Please forgive my misunderstanding. However, if PPLB
is NOT widely used, why would you particularly care
about its effects? Avian Carriers are not widely used
either, and I don't much care about their effect on

Second, the router architecture issue about whether
PPLB was possible on
certain routers. It is possible on a great number of
routers. But there
are some details I missed.

Here I disagree: you made statements about the default
behavior of Cisco and Juniper routers which reflected
an incorrect understanding of the actual workings and
deployed configurations of same. My argument that
strenuous assertions of incorrect facts weakens
credibility holds.

Please don't put (wrong) words in my mouth, and then
say I'm wrong.

I apologize if I misquote or distort in any way, it is
certainly not my intent. Any search of my previous
postings to NANOG would show that I attempt to be
accurate in representing and commenting on others'

David Barak
Need Geek Rock? Try The Franchise:

Yada Yada Yada - PPLB - Yada Yada Yada.

Hey, enough already. Cut it out. Trolls never get full.

Listen, by now we all know that PPLB does not behave well with anycast.
Those who actually have networks that use PPLB have been warned. Repeatedly.
So I am sure they'll be appropriately cautious, and they appreciate all the
advice and education and being made aware of the issues, if they didn't
already know. While barely "real-time operational" (I don't recall anyone
waving a red flag and saying that their network was broken right now from
this), the subject is done.

I don't run PPLB in my network. So I don't give a damn beyond being aware
that down the road if I have to trouble shoot issues, PPLB will be in the
back of my mind. Thanks for that. I don't know how many do run PPLB. And
which of them run anycast behind it. I don't care about that either. They
run their networks. I run mine. If I want to learn about the intricacies,
I'll go look for the appropriate venue or forum. It shouldn't be NANOG.

But this incessant noise complete with repetition of the noise is *not*
appropriate for the list by any sane definition. If it is now considered
appropriate, then NANOG is no longer an operational mailing list, and I'm
going to look for - or start - another one.

Suggesting the use of procmail recipes to filter the latest reincarnation of
the "kooks korner" is wrong. Life's too short, and I don't have time.

As I said earlier this year (about NANOG) in another forum:

" I have an interest in an organization that can act as a point of
"attraction" for clueful geeks who run networks that carry real traffic that
my packets interact with on a constant basis, and where any issue that they
have with their networks often affects my ability to make my bits go where
they are intended, and vice versa. I don't mind it being populated by
clueful geeks from equipment vendors who can contribute to operational
discussions by providing information or advice that relates to their
products or the operation of networks in general, without smothering me with
sales puffery. And who can also take information away from such discussion,
and make improvements in their products to help the network I play on

I want people who actually do things on networks, rather than people who
pontificate about what aught to happen on networks. I want people who can
practice Gestalt Protocol without bullshit. People who can say "In my
experience..." and actually have some that is relevant. I want people who
can say "This is what you're seeing, and this is what is causing it, and
this is how to fix it", and it does. And I want people who can say "It's my
responsibility, I'll jump on it and fix it", and who will."

I don't know where nanog-futures has gone with deciding on nanog-futures. I
hope that the "carte blanche" that seems to have been granted to everyone
with their threads and attacks and pontification has been done in order to
show the "members" how bad it can get. And that it is going to stop. Because
if it is not, I'm outta here, after almost 11 years. Some of the stuff on
the list over the last 2 months is downright embarrassing to be associated
with. It seems to me that while some are here to actually improve the state
of the Internet, far too many are here to improve their positions in their
own alternative universes, and minds. And that's sad.

There, that's my one email for today. Or longer.

Rodney Joffe
CenterGate Research Group, LLC
"Technology so advanced, even WE don't understand it"(R)