[nycbug-talk] IPv6 and CIDR confusion

[Jesse Callaway]

On 3/23/07, Isaac Levy <ike at lesmuug.org> wrote:
> On Mar 23, 2007, at 6:01 PM, alex at pilosoft.com wrote:
>
> > On Fri, 23 Mar 2007, Isaac Levy wrote:
> >
> >> Hi All,
> >>
> >> When searching for info about IPv6 netblocks, I found: "IPv6 uses a
> >> CIDR-style architecture for address allocation"
> >>
> >> -but it's not actually CIDR?
> > It is classless.
>
> Gotig- learning something new every day.
>
> >
> >> Does anyone know where to find out about the breakdown of netblocks?
> >> (e.g. what is a /64 for IPv6?  What is the equivalent of an IPv4
> >> Class
> >> C, (/24 with 254 usable addresses?))
> > There's no such thing as classes. Duh.
> >
> > /64 for ipv6 is still the same, SLASH SIXTY FOUR.
>
> Sure, but it doesn't seem to be the same...
>
> > On Mar 23, 2007, at 5:00 PM, Jonathan Stewart wrote:
> >> I was hoping to find something on wikipedia that would be a bit
> >> easier to work through for someone new to it but the RFC always
> >> works :)
> >> http://tools.ietf.org/html/rfc2373
>
> - so,
>
> The IPv4 CIDR number comes from the number of 1's in the subnet mask
> when converted to binary, right?
>
> So it seems the IPv6 notation is not the same thing at all- it's not
> the subnet mask, but the prefix length of the address.
>
> That was confusing, now it's clear.
>
> >
> > To have 254 usable IP addresses, you can have a /120.
>
> Alex- do you do the hexadecimal counting in your head dude?!?!?
>
> >
> > Now, the interesting thing is, most v6 capable routers only will
> > route on
> > the top 64 bits of the prefix.
>
> - whaddya' mean?  That sounds interesting?
>
> Rocket-
> .ike
>
> (p.s.: for a guy who seems so hellbent against IPv6, you sure know a
> lot of details...)
>
>

It means that if you get more specific about your netmask the router
won't be able to differentiate and send it on a different route.

-jesse