[address-policy-wg] Re: Fallacy by Kurt (was Re: IPv6 Policy Clarification - Initial allocation criteria "d)")

On Tue, 22 Jun 2004 22:41:20 +0900, Masataka Ohta wrote:
>Do you know that the current Internet backbone is operating
>with parallel 10Gs?
Someone may stretch the term "backbone".
There is no single one.
Some use OC-3 and some use OC-192, some plan to use
40G (and caclculate wavelength counts against bandwidth
occupied by a single wavelength).

>So, large memory costs.
Not really.

Todays price, single pcs., IDT/-100 at Arrow for 64K*72 TCAM:
$155 from stock.

If someone operates a major backbone, this is no real
cost factor compared to leased lines and wavelengths ...

Even if you multiply this with five for sales droid financing ...

Fast synchronous static memory is even *much* less, e.g.
few $'s per Multi-Mbit chip.

>Do you also know that access speed of memory (including but
>not limited to TCAM) degrades proportional to log or sqrt of
>the number of entries?
Again, it depends ...

>Both modern routers and modern CPUs are highly pipelined, which
>means there is some performance loss if TCAM or primary cache
>miss occurs.
A *cached* router is a good-weather-only product which would
die e.g. if a DDoS or SQL Slammer is on the road.

Thus modern backbone routers *do not use route caches*.

>Secondary or third level cache of modern CPUS often have millions
>of entries and constructed with static RAM.
Which again tells us that 10K is not a real limit for modern
routers, as well as 100K is not.
Noone would buy a *new* backbone router which isn't capable
to handle a n*100K table even if there is "rain and snow" inside
the network ...

Best Regards
Oliver Bartels



Oliver Bartels F+E + Bartels System GmbH + 85435 Erding, Germany
oliver at bartels.de + http://www.bartels.de + Tel. +49-8122-9729-0