[34981] in North American Network Operators' Group
Re: [NANOG] RE: rfc 1918?
daemon@ATHENA.MIT.EDU (Pim van Riezen)
Thu Feb 22 20:03:59 2001
In-Reply-To: <20010223004819.DD65F8B@proven.weird.com>
Date: Fri, 23 Feb 2001 01:57:08 CEST
From: Pim van Riezen <pi@vuurwerk.nl>
To: nanog@merit.edu (North America Network Operators Group Mailing List)
Message-ID: <PstOfc.3a95b564.002f1c@localhost>
Errors-To: owner-nanog-outgoing@merit.edu
woods@weird.com (Greg A. Woods) tapped some keys and produced:
> In practical terms I suppose it also depends on just exactly what
> filtering technology you've deployed, and just exactly how close it is
> to being overloaded. If you are already pushing your router's CPU too
> hard (and if your filters are done by your router's CPU rather than an
> ASIC) then obviously reducing your filter load will be in your own best
> interests and not filtering destination addresses against RFC-1918 will
> be one relatively benign way of reducing the filter load. However if
> your router's CPU is only partially utilised now (even if you push your
> pipe to capacity), then adding such destination filters won't hurt
> anyone.
Would routing them to Null0 not be more optimal?
Pi
--
conf t
no ip-directed marketing drivel
^Z
wr mem
