Hello Eliot
In addition to what you cited below from Geoff's research, there is an
other study suggesting that the advances in multi-core processors and
memory technology are "sufficient to keep the sky from falling".
Routing Tables: Is Smaller Really Much Better? Kevin Fall. Gianluca
Iannaccone. Sylvia Ratnasamy. Intel Labs Berkeley. P. Brighten Godfrey
University of Illinois at Urbana-Champaign:
http://berkeley.intel-research.net/sylvia/hotnets2009-final156.pdf
Published in Hotnets2009.
Best regards
Hannu
________________________________
From: [email protected] [mailto:[email protected]] On
Behalf Of ext Eliot Lear
Sent: Tuesday, May 25, 2010 12:41
To: Tony Li
Cc: IRTF Routing RG; RJ Atkinson
Subject: [rrg] moving toward the next straw poll for q 1?
Tony, Ran, All,
First, thanks very much to Ran for conducting the poll and
generating a substantial amount of consensus of view. Tony has asked
for additional discussion on (1). For everyone's information here is
the text:
(1) "The Internet continuing down the current
architectural path, whereby site multi-homing increases the size/entropy
of the DFZ RIB/FIB, is not believed to be scalable or viable."
As one of the people who withheld my vote on this point, I will
explain why I did so. My own view is that the wording was a bit open
ended, and I would like to have seen it qualified just a bit more. It
seems to me that Geoff's latest research is worth keeping in mind.
There he pointed out that the number of updates has been constant over
some period of time, even as the number of prefixes is increasing.
We have a number of unknowns in the medium term:
* What will be the impact of IPv6 transition/coexistence?
* What will be the impact of IPv4 exhaustion (this is
perhaps the other side of the same coin)?
* How will networks connect to each other in the future?
For example, will the southern hemisphere become considerably more
interconnected than it is now?
* How will/should consumers multihome?
If we cannot answer these questions we can certainly assert that
these are risks relating to the scalability of the routing system.
You'll note I'm shying away from mobility for the moment, but
that may be an elephant in the room.
Regards,
Eliot
On 5/25/10 2:34 AM, Tony Li wrote:
Hi all,
It's now considerably after the official close of this
poll.
As of this writing, there were 33 responses, of which 1
is clearly invalid.
That leaves us with 32 valid responses.
Statement 1 has (20/32) 65.6% agreement.
Statement 2 has (26/32) 81% agreement.
Statement 3 has (32/32) 100% agreement.
Statement 4 has (31/32) 97% agreement.
All of these clearly represent rough consensus.
However, I would be happier
if Statement 1 had a slightly more decisive consensus.
Therefore, I'd like
to ask the group to continue to discuss, refine, and
repeat the poll on a
revised version of Statement 1. I suspect that simple
wordsmithing will
suffice.
The other 3 statements will show up in the next rev of
the document.
Regards,
Tony
On 5/24/10 10:58 AM, "RJ Atkinson" <[email protected]>
<mailto:[email protected]> wrote:
The poll closes in just under 1 hour (from when
I write this note).
If anyone has not yet participated and would
like to do so,
PLEASE go do so RIGHT NOW. As a reminder, the
poll URL remains:
<http://www.doodle.com/z5s9yq8kt73eua9t>
<http://www.doodle.com/z5s9yq8kt73eua9t>
As I write this, there are 31 valid
participants, and one
spoilt ballot. Votes are public, visible via
the above URL.
For the record, no one has been censored,
although someone
failed in an attempt at humour by entering a
spoilt ballot
which s/he herself/himself labelled "Censored
by...".
Under IRTF process rules, determination of RG
consensus is
the privilege of the RG chairs.
So, purely as an observation, not a consensus
determination,
most current (13:52 EDT, 24th May 2010)
participants support
all 4 statements.
Statement 1 currently has (19/31 or ~61%)
agreed.
Statement 2 currently has (25/31 or ~81%)
agreed.
Statement 3 currently has (31/31 or 100%)
agreed.
Statement 4 currently has (30/31 or ~97%)
agreed.
Personally, I am encouraged by all this.
Yours,
Ran
_______________________________________________
rrg mailing list
[email protected]
http://www.irtf.org/mailman/listinfo/rrg
_______________________________________________
rrg mailing list
[email protected]
http://www.irtf.org/mailman/listinfo/rrg
_______________________________________________
rrg mailing list
[email protected]
http://www.irtf.org/mailman/listinfo/rrg
