Re: [multimob] MLD HLD Message

>   We had a discussion on where MLD/IGMP Hold message needs to be
> specified. There are two options: either it is specified in MLD/IGMP
> Mobile draft or in individual protocol extension draft(s) for
> HMIP/MIP, etc.
> 
>   Please post your opinions.

Maybe I should clarify.

I sent a mail to the multimob ML on Apr.14.

Thomas proposed MLD hold message in;
http://tools.ietf.org/html/draft-schmidt-waehlisch-mhmipv6

It assumes the use of MLD hold message with HMIP6.
On the other hand, MLD hold message might be useful for fast handover
scenario in general, because MLD hold asks to the upstream mrouter or
proxy (i.e. MAPs or HA) to keep join state during MN's movement and
recover the multicast session after MN's movement.

On the other hand, one may think that MLD hold state is not mandatory
as the general MLD extension, because the fast handover would be much
faster than the time of membership expiration maintained by MLD. This
means even if MN does not send any MLD message to his upstream router
or proxy, he can recover the multicast session quickly since he can
move to the new mobile network very fast (i.e. faster than MLD
expiration time).
I don't know if it's always true or not.

So now, I'd like to hear your opinions.
If MLD hold is useful especially (or only) for HMIP, then MLD hold
specification would be kept in the above HMIP draft.
If it is useful to propose it as the MLD extension as the general
function, I'm happy to work for defining the specification in the MLD
extension draft with Thomas.

Thank you for your input.
--
Hitoshi Asaeda
_______________________________________________
multimob mailing list
multimob@ietf.org
https://www.ietf.org/mailman/listinfo/multimob