Re: [Mip4] RFC3344biz Dynamic HoA bug ?

[Kent Leung <[e-mail]>]

Yes, this is a known problem for many years.  :)  In addition, 
draft-mccann-mobileip-sessionid 's
Address Management section attempted to covered this issue as well.

We haven't addressed this since most SP deployments have not had the need to
return to home network.  And for Enterprise environment where DHCP is 
prevalent,
Steve Glass's draft-glass-mobileip-agent-dhcp-proxy made sense.

I agree with Hasse that 3344bis should just state that when MN returns 
home, the dynamically
assigned address needs to remain associated with the MN.  The details to be 
covered in another
draft.

Kent

At 09:19 AM 3/25/2004 +0200, Bjorn Andersson wrote:
>On Wed, Mar 24 2004, at 21:03:26 -0500, O'Neill Alan wrote:
> > A MN places 0.0.0.0 into the Home Address field of the RREQ to trigger
> > assignment by the Home Agent. The address allocation lifetime is assumed to
> > be the MIP binding lifetime because no other information exists, and
> > permanent allocations are of course not good. However, what then 
> happens when
> > the MN returns home ?
>
>This has been discussed before on this list and I think that this is
>really a problem that we should solve before we have n proprietery
>solutions. Good that you brought up the discussion again! There
>have been at least two drafts that have been addressing this
>problem. Without any further digging I found these:
>- draft-glass-mobileip-agent-dhcp-proxy
>- draft-thuel-mobileip-tt
>
> > The MN cancels the binding, and according to the spec ceases to use MIP
> > signaling. However, if the address lifetime is tied to the binding lifetime
> > then the MN just lost its address. I assume this might not have been 
> spotted
> > before because we are mostly focusing on cellular systems where the MN 
> has a
> > virtual home network, and hence cannot ever get home, but given the 
> scope of
> > RFC3344 this 'hole', if people concur, needs to be filled. The following
> > options spring to mind.
>
>I have been focusing on non-cellular systems for a number of years and
>come across this problem several times in real deployment scenarios. So
>far this can only be solved with proprietery solutions and I really
>support standardizing a solution. Something along the lines with your
>proposed HALG/HALR sounds reasonable. I think this is the way to go.
>
>   Bjorn
>
>--
>Bjorn Andersson <[e-mail]>
>PGP id 5AFC144B
>
>--
>Mip4 mailing list: [e-mail]
>     Web interface: https://www.ietf.org/mailman/listinfo/mip4
>      Charter page: http://www.ietf.org/html.charters/mip4-charter.html
>Supplemental site: http://www.mip4.org/


-- 
Mip4 mailing list: [e-mail]
    Web interface: https://www.ietf.org/mailman/listinfo/mip4
     Charter page: http://www.ietf.org/html.charters/mip4-charter.html
Supplemental site: http://www.mip4.org/