Home
Vendor
Project
  Phase I
Press
  Phase I
About
You Are Using IPv4 To Access This Site
Your IP Address is: 38.107.191.82

RE: [moonv6] /120 prefix length at UNH

New Message Reply Date view Thread view Subject view Attachment view

From: Bound, Jim (jim.bound@hp.com)
Date: 10/16/03


moonv6 post from "Bound, Jim" <jim.bound@hp.com> Ed,

Not quite RFC 2461 and RFC 2462 are agnostic to the address prefix 000 or 001.

>From ND RFC 2461 Prefix option for RA:

Prefix Length 8-bit unsigned integer. The number of leading bits 

                     in the Prefix that are valid.  The value ranges
                     from 0 to 128.

Also EUI-64 is ONLY a definition of the Interface Identifier for non 000 prefix.

We are moving to /64's that is fine. Not an issue for me.

But this is important discussion. What I am concerned now about is that implementations may have hard coded a 64 bit prefix and 64 bit EUI boundary in their code. That is clearly wrong per the IPv6 architecture you must assume search on the entire 128bits.

2461 and 2462 MUST be agnostic to what your saying. And they are.

Lets be careful here folks. This is pretty basic technical stuff.

Also the prefix length announced as RA for a link can be different that the prefix length used for routing packets. We were speaking of the latter not the former. Two different discussions.

I am offline for a few days no other option.

Thanks
/jim

> -----Original Message-----
> From: Ed Remmell [mailto:eremmell@elmic.com]
> Sent: Wednesday, October 15, 2003 10:41 PM
> To: Bound, Jim; 'Alain Durand'
> Cc: schultz@io.iol.unh.edu; moonv6@iol.unh.edu
> Subject: RE: [moonv6] /120 prefix length at UNH
>
>
> Hosts are still restricted to using EUI-64 interface IDs, w/
> regards to Stateless Address Auto-Configuration. So, a host
> cannot auto-configure from a /120 prefix. You can advertise a
> /120 prefix in a RA Prefix Information Option, but that
> doesn't make too much sense to do so, since any host
> compliant with the RFCs will not be able to configure from
> such a prefix. If you want to instead use DHCPv6 or manual
> address configuration, then it doesn't really matter, you can
> choose the prefix length to be as long as you'd like in that
> case IMHO.
>
> Thanks.
> - Ed
>
> > -----Original Message-----
> > From: owner-moonv6@io.iol.unh.edu
> > [mailto:owner-moonv6@io.iol.unh.edu] On Behalf Of Bound, Jim
> > Sent: Wednesday, October 15, 2003 10:09 AM
> > To: Alain Durand
> > Cc: schultz@io.iol.unh.edu; moonv6@iol.unh.edu
> > Subject: RE: [moonv6] /120 prefix length at UNH
> >
> >
> > moonv6 post from "Bound, Jim" <jim.bound@hp.com>
> > Good response.
> >
> > OK I would suggest the EUI can be part of the prefix
> > ......................that does not break 3315. The only
> > thing that breaks 3315 is if there is no way to extrapolate
> > the EUI a routing prefix does not preclude that is where I am????
> >
> > /jim
> >
> > > -----Original Message-----
> > > From: Alain Durand [mailto:Alain.Durand@Sun.COM]
> > > Sent: Wednesday, October 15, 2003 1:05 PM
> > > To: Bound, Jim
> > > Cc: schultz@io.iol.unh.edu; moonv6@iol.unh.edu
> > > Subject: Re: [moonv6] /120 prefix length at UNH
> > >
> > >
> > >
> > >
> > > Bound, Jim wrote:
> > >
> > > >With /120
> > > >
> > > >m = 8. n == variable depending on prefix format. this is
> > > aggregatable
> > > >format. so the prefix stops at /64 and low order 64 is the
> > > EUI. meaning
> > > >n MUST not be greater than 58. 58 + 8 == 64. So what I am
> > > saying is I
> > > >beleive /120 or /96 does not step on the EUI but a /8 does
> > > step on the
> > > >EU and does break 3315?
> > > >
> > >
> > > | n bits | m bits | 128-n-m
> > bits |
> > >
> > +------------------------+-----------+----------------------------+
> > > | global routing prefix | subnet ID | interface
> > ID |
> > >
> > +------------------------+-----------+----------------------------+
> > >
> > > I think you might have been loonking at the picture right to left
> > > instead of left to right! (just kidding)
> > >
> > > With /120, you have:
> > >
> > > | 120 bits |
> > 8 bits |
> > >
> > +------------------------+-----------+----------------------------+
> > > | global routing prefix + subnet ID
> > |interface ID|
> > >
> > +------------------------+-----------+----------------------------+
> > >
> > >
> > > Ok? So your Interface ID is effectively 8 bit long. This
> breaks 3315
> > > as the global routing prefix is not starting with 000/3.
> > >
> > > - Alain.
> > >
> > >
> >
> > ---
> > Incoming mail is certified Virus Free.
> > Checked by AVG anti-virus system (http://www.grisoft.com).
> > Version: 6.0.522 / Virus Database: 320 - Release Date: 9/29/2003
> >
> >
>
> ---
> Outgoing mail is certified Virus Free.
> Checked by AVG anti-virus system (http://www.grisoft.com).
> Version: 6.0.522 / Virus Database: 320 - Release Date: 9/29/2003
>
>
>

New Message Reply Date view Thread view Subject view Attachment view

This archive was generated by hypermail 2.1.7 : 12/01/06 EST