In your previous mail you wrote:
We are not solving a link layer problem here.
=> I disagree as the dormant mode support is at the link layer.
Problems are the short interval between periodic multicast
^^^^^
RAs and an MN currently does not solicit for RAs rather depends on
periodic RS to update the default router lifetime.
=> the interval should not be considered as (too) short without the
dormant mode, so the issue is the dormant mode and not the interval.
Nonetheless, you can argue that the current specified values for
these are good enough, which can be debated.
=> this is exactly I'd like to argue!
> => IMHO the best idea to solve this class of issues is in
> draft-ietf-dna-frd-01.txt (i.e., cache the last RA in the AP).
I am one of the contributors for this draft; and I think FRD
is unrelated to our context. FRD expedite the RA acquisition
and does not deal with periodic multicast RAs.
=> the solution I propose is to cache the last RA in the AP and
to send it to the node when it wakes up. This is place to wake up
it for each RA when it sleeps. So this is not exactly FRD but
it is very close.
Link layers are already dealing with dormant mode, but network layer
may be disturbing this mode; so the draft is proposing few changes to
the network layer.
=> what I strongly suggest is to fix the dormant mode instead.
Regards
[EMAIL PROTECTED]
PS for Basavaraj: direct messages are blocked by a stupid RBL issue,
have you an alternate e-mail address?
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
