Re: [Mipshop] FMIPv6 Reactive Handover - HI/HAck

[Rajeev Koodli <[e-mail]>]

Narayanan Vidya-CVN065 wrote:

> 
>     Hi Subba Reddy,
>            
>     --> As per my understanding, FBack is sent to MN's NCoA (sec 6.3.2).
>     But in the figure, it is showing to NAR.  
>      

Actually, the Figure only shows FBack being sent _towards_ NAR. The 
point being it is forwarded, just as any other packet, once NAR
detects MN's attachment. NAR does not know, and it must not be required 
to, which packet contains FBack.


>     ------> It does seem that the FBAck is sent to the NCoA (as per the
>     IP destination value in 6.3.2, although this doesn't make much sense
>     if it is sent on the old link itself, as in the predictive handoff

You have to go back to archives for this sort of thing :-)
(Briefly, the MN is not required to stay on the old link after
sending FBU. Hence FBack is always sent to NCoA. But if it happens
to be hanging around on the previous link, it can still receive one
if PAR sends it on the previous link)

-Rajeev


>     case). Nevertheless, we could take advantage of the fact that it is
>     going through the NAR for sure in the case of reactive handovers.
>      
>     Vidya
>      
>      
>     Thanks & Regards,
>     Subba Reddy
> 
>          
>         Again, CT is a different story - so, I'm not referring to that.
>          
>         I think the text can be made clearer in the protocol description
>         to alleviate ambiguity. But, regardless, why is there a need to
>         mandate these messages for reactive mode?
>          
>         I can see the need for a message exchange with the AH between
>         the ARs - but, seems to me like the FBU/FBAck can be extended to
>         carry that - no?
>          
>         Vidya
> 
>             -----Original Message-----
>             From: Junghoon Jee [mailto:[e-mail]]
>             Sent: Monday, August 22, 2005 9:15 PM
>             To: Narayanan Vidya-CVN065; [e-mail]
>             Subject: RE: [Mipshop] FMIPv6 Reactive Handover - HI/HAck
> 
>             Hi Vidya,
>             I made the same question through DNA WG recently regarding
>             FMIP-DNA I-D.
>              
>             You can find the previous discussion through the attached
>             message.
>              
>             I am still questioning why HI/HAck for the access control is
>             required after MN has already attached to NAR in the
>             reactive case.
>              
>             Junghoon
>              
>>  -----Original Message-----
>>  From: Rajeev Koodli [mailto:[e-mail]]
>>  Sent: Tuesday, July 19, 2005 8:46 AM
>>  To: Junghoon Jee
>>  Cc: 'Syam Madanapalli'; [e-mail]
>             <mailto:[e-mail]>; [e-mail]
>             <mailto:[e-mail]>
>>  Subject: Re: [DNA] draft-koodli-dna-fmip-00.txt
>>
>>
>>
>>
>>  Junghoon Jee wrote:
>>
>>  > Rajeev,
>>  > 
>>  >
>>  >>For both predictive and reactive modes, HI/HAck can be
>             useful for
>>  >>access control. NAR can forward packets for the MN
>             without forcing
>>  >>access control, through HI/HAck.
>>  >
>>
>>  One could use HI/HAck for CT, but I am not referring to that.
>>
>>  When FBU processing is successful, PAR's message (HI) to NAR
>>  can allow NAR to "validate" the ND cache entry for NCoA.
>>  No Context Transfer is involved. The fields in HI (PCoA, NCoA,
>>  LLA) are sufficient for the purpose.
>>
>>  -Rajeev
>>
>>
>>
>>
>>
>>  >
>>  > Are you saying about context transfer by HI/HAck ?
>>  >
>>  > Regards,
>>  > Junghoon
>>  >
>>  >
>>  >>(If it is not clear, FMIP-DNA is applicable when
>             DNA+oDAD is
>>  >>available,
>>  >>  and no neighborhood information is available for FMIP)
>>  >>
>>  >>-Rajeev
>>  >>
>>  >>
>>  >>
>>  >>Syam Madanapalli wrote:
>>  >>
>>  >>
>>  >>>Hello Junghoon,
>>  >>>
>>  >>>Thanks for reviewing the draft.
>>  >>>
>>  >>>
>>  >>>>Hi Syam,
>>  >>>>It's an interesting work.
>>  >>>>After reviewing this I-D, I've come up with a
>             following question.
>>  >>>>
>>  >>>>About the role of HI/HACK in this reactive FMIPv6-DNA :
>>  >>>>I thought that these messages are required to confirm
>             the MN's
>>  >>>>prospective NCoA in predictive mode.
>>  >>>>In the FMIPv6-DNA, MN configures the NCoA by optimistic
>>  DAD in the
>>  >>>>reactive mode, so why do those messages needed to be
>>  >>
>>  >>transferred and
>>  >>
>>  >>>>what's their roles ?
>>  >>>>Just to prepare for future optional behavior the same as
>>  >>
>>  >>specified in
>>  >>
>>  >>>>the draft-ietf-mipshop-fast-mipv6-03 ?
>>  >>>>I guess that my question may also apply to the
>             reactive mode of
>>  >>>>draft-ietf-mipshop-fast-mipv6-03.
>>  >>>
>>  >>>
>>  >>>As you mentioned, currently there is no new use of HI/HACK
>>  >>
>>  >>other than
>>  >>
>>  >>>what is mentioned in FMIPv6 RFC.  That is, NAR can make
>             use of the
>>  >>>knowledge that its trusted peer (i.e., PAR) has a trust
>>  >>
>>  >>relationship
>>  >>
>>  >>>with the
>>  >>>MN.
>>  >>>
>>  >>>
>>  >>>
>>  >>>>Regards,
>>  >>>>Junghoon
>>  >>>>
>>  >>>
>>  >
>>  
> 
>         ------------------------------------------------------------------------
> 
>         _______________________________________________
>         Mipshop mailing list
>         [e-mail]
>         https://www1.ietf.org/mailman/listinfo/mipshop
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Mipshop mailing list
> [e-mail]
> https://www1.ietf.org/mailman/listinfo/mipshop


_______________________________________________
Mipshop mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/mipshop