In a message written on Fri, Jan 22, 2010 at 07:09:00PM +0000, Nick Hilliard wrote:
What portion of 1/8 is untainted? Or any other /8 that the IANA has identified as having problems? How do you measure it? How do you ensure
I, personally, am quite skeptical that any of the /8's are tainted to the point where they are unusable. However, as an example of something I would say rises to the level of tainted. Remember a few years back Netgear hard coded the IP's of the UW time servers, and then shipped a few million boxes, which on by the way had a bug so they asked for time too often? http://pages.cs.wisc.edu/~plonka/netgear-sntp/ The result was a 40,000 packet per second flood. If an RIR was to give out a block with that sort of taint (e.g. UW returned that block, or something out there defaults to contacting 1/8 in a similar manor) then I think it's reasonable for the RIR to mark it as tainted, work with the people to get it fixed, and give folks other address space. Hopefully the RIR could work with the party involved and eventually return the space to service.
IANA hands out /8s. We know that some of these are going to cause serious problems, but life sucks and we just have to deal with what happens.
Pretty much. -- Leo Bicknell - bicknell@ufp.org - CCIE 3440 PGP keys at http://www.ufp.org/~bicknell/