[31135] in North American Network Operators' Group

home help back first fref pref prev next nref lref last post

Re: When IPv6 ... if ever?

daemon@ATHENA.MIT.EDU (Masataka Ohta)
Wed Sep 13 09:19:01 2000

From: Masataka Ohta <mohta@necom830.hpcl.titech.ac.jp>
Message-Id: <200009131310.WAA12705@necom830.hpcl.titech.ac.jp>
In-Reply-To: <Pine.BSF.4.21.0009121327340.21511-100000@shell.inch.com> from Charles
 Sprickman at "Sep 12, 2000 01:36:50 pm"
To: Charles Sprickman <spork@inch.com>
Date: Wed, 13 Sep 2000 22:10:25 +0859 ()
Cc: batz <batsy@vapour.net>, "Roeland M.J. Meyer" <rmeyer@MHSC.com>,
	nanog@merit.edu
Errors-To: owner-nanog-outgoing@merit.edu


Charles and others;

Thank you very much for the information.

> > As for the performance issue, can someone give me a pointer
> > to *INEXPENSIVE* NAT box which can operate at 100Mbps, or better,
> > 1Gbps?
> 
> Inexpensive can be a pretty loose term.

Oops, sorry.

As you might know, with Ethernet, the cost of a 100Mbps OE is about $200,
pair of which will be used for each subscriber. (I'm overestimating the
cost of OE by not assuming a dedicated hardware)

So, a 100Mbps NAT box should costs $100 or, maybe, $200.

Actually, 100Mbps case is not very realistic, because of the management
cost.

A 1Gbps NAT box may be shared by 100 subscribers and may cost $10000
or, maybe, $20000.

> > I was asked by ISPs which are beginning to offer 100Mbps Ethernet service
> > to subscribers.
> 
> They probably have a little cash then...

They have little cash to be spent on equipments provided for each
subscriber and replaced two years later.

> > My guess is that there is none.
> 
> This product, while primarily a firewall, demonstrates that doing NAT and
> firewalling (lots of shared code there, see IPFilter) on an ASIC makes for
> a much cheaper solution than FW-1 on a big Sun Ultra-something...

Hmmmm, $109,000 seems to be too expensive.

> Anyhow they claim lots of sessions:

Each subscriber won't need so many sessions.

							Masataka Ohta



home help back first fref pref prev next nref lref last post