What you say is correct of course, but it does not seem to be what current implementations actually do.
RFC 2462 says "Duplicate Address Detection is performed on unicast addresses prior to assigning them to an interface..." If a node returns to an already visited network for which it has defended an address, does using it again consist of "assigning" it to the interface? As you say there is a case why it should, but current implementations, at least those with which we are familiar, don't work that way. Richard. ----- Original Message ----- From: "Alper E. YEGIN" <[EMAIL PROTECTED]> To: "Richard Nelson" <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]> Cc: "Thomas Narten" <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]> Sent: Saturday, October 12, 2002 10:11 AM Subject: Re: Changing RS Reply Timing for Mobile IPv6 > > > > The short answer is that that is what Linux/MIPL does. We haven't changed > > anything. > > The slightly longer answer is that RFCs2461/2462 describe how prefixes are > > advertised with a valid lifetime > > and that addresses take on that lifetime and can be used until that > lifetime > > expires. RFC2461 specifically describes the case where a laptop is > > unplugged and is entitled to use its previous address when plugged in > again. > > The default value for lifetimes is 30 days and infinity is a valid value. > > This lifetime is about "prefix validity". Not "ownership > of an IP address configured based on this prefix". > > Any node can attempt to configure any IP address > with a valid prefix. Unless there is another node > defending this address, DAD succeeds, hence IP > address is configured. > > In the case of your mobile node being away for a while, > anyone else can claim it's care-of address and configure > it in the absence of your node defending it. > > (of course, if there is a home agent on that link, with > this IP address in its binding cache as the home address > of a mobile node, then this home agent can defend the address. > this would be the case if your mobile node is using forwarding > from previous care-of address technique when it moved > to the other visited network) > > alper > > > > > > > > Richard. > > > > ----- Original Message ----- > > From: "Alper E. YEGIN" <[EMAIL PROTECTED]> > > To: <[EMAIL PROTECTED]>; "Richard Nelson" > > <[EMAIL PROTECTED]> > > Cc: "Thomas Narten" <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]> > > Sent: Saturday, October 12, 2002 4:15 AM > > Subject: Re: Changing RS Reply Timing for Mobile IPv6 > > > > > > > Greg, > > > > > > > I'd just like to indicate that we have seen the delay > > > > for router advertisement to be a significant delay > > > > in MIPv6 handovers, since there is no DAD in the case > > > > where a mobile node moves back to a previously visited > > > > network. > > > > > > How is that so? Of course the chances of some other > > > node claiming your node's IP address while it's gone > > > are even less, but still it is not impossible. So, you can > > > choose to not use DAD by taking this risk... Is this what > > > you mean? > > > > > > alper > > > > > > > > > > > > -------------------------------------------------------------------- IETF IPng Working Group Mailing List IPng Home Page: http://playground.sun.com/ipng FTP archive: ftp://playground.sun.com/pub/ipng Direct all administrative requests to [EMAIL PROTECTED] --------------------------------------------------------------------
