Re: [DNA] [Issue 15] [Issue 16] DAD and MLD Interaction

[James Kempf <kempf@docomolabs-usa.com>]

Erik,

Yes, I agree, I think we discussed this in the DT. But I think it's still
possible that it might be an error case.

There two error cases are that the host receives a multicast RA w.o. any
Link Up, and the host receives an unsolicited unicast RA w.o. a any Link Up.
In a separate email, I proposed a couple ways to handle this, maybe you
could take a look at it and comment?

It's also possible that the host receives a multicast RA with Link Up, as,
for example, if it sends an RS and the router is overloaded, but that case
should be covered by the procedure for what to do before sending a DNA RS.

            jak

----- Original Message ----- 
From: "Erik Nordmark" <erik.nordmark@sun.com>
To: <greg.daley@eng.monash.edu.au>
Cc: "James Kempf" <kempf@docomolabs-usa.com>; <dna@eng.monash.edu.au>
Sent: Wednesday, June 22, 2005 4:44 AM
Subject: Re: [DNA] [Issue 15] [Issue 16] DAD and MLD Interaction


> Greg Daley wrote:
>
> > Should this be tied to the Link-Up indications, or to the DNA reception
> > and processing of hints?
> >
> > For example, if no link-layer indication is received does the
> > reception of an 'unexpected' RA and subsequent initiation of
> > DNA operations cause Optimistic to be set?
> >
> > I'd guess it's tied to DNA initiation, rather than the Link-Up
> > explicitly.
>
> Things wouldn't work safely from a DAD perspective if there is no
> link-up notification.
>
> Take an example:
> 1. Host moves and attaches to link B (no link-up notification)
> 2. Host receives a NS for it's link-local address. It responds since it
>     isn't aware that it might have moved to a different link.
> 3. Host receives an unsolicited RA, which has some 'unexpected' things
>     in it (e.g., only unknown prefixes). This triggers DNA; move
>     addresses to optimistic state and send RS.
>
> The issue is that if the link-local address is a duplicate on the new
> link, then at #2 the host would mess up things for the existing user of
> that link-local address on the new link.
>
> So things can't be very robust without a link-up notification.
>
>     Erik
>
>
>