[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: RE: questions about draft-wen-ipv6-rsra-opt-multihoming-00



Hi, haibo,
 I think maybe it is more meaningful to discuss how to select the source
address for the node in the multihoming site.

 Best Regards, Sam Xia 

> -----Original Message-----
> From: owner-shim6@psg.com [mailto:owner-shim6@psg.com] On 
> Behalf Of CTO WEN Haibo
> Sent: Friday, June 09, 2006 1:12 PM
> To: Geoff Huston; Greg Daley
> Cc: Lawrence Zou; shim6@psg.com
> Subject: RE: RE: questions about 
> draft-wen-ipv6-rsra-opt-multihoming-00
> 
> Yes, I think it's relevant to the shim6 framework.
> 
> "Shim6 working group will produce specifications for an 
> IPv6-based site multi-homing solution that inserts a new 
> sub-layer (shim) into the IP stack of end-system hosts. It 
> will enable hosts on multi-homed sites to use a set of 
> provider-assigned IP address prefixes and switch between them 
> without upsetting transport protocols or applications."
> 
> With the help of multi-homing information option along with 
> the prefix information option, shim6 layer can directly find 
> the matching source address to reach the services from the 
> desired ISP. Of course, shim6 layer can further try other 
> source address to reach the service if needed.
> 
> Haibo
> 
> > -----Original Message-----
> > From: Geoff Huston [mailto:gih@apnic.net]
> > Sent: 2006?6?9? 09:00
> > To: CTO WEN Haibo; Greg Daley
> > Cc: Lawrence Zou; shim6@psg.com
> > Subject: RE: RE: questions about
> > draft-wen-ipv6-rsra-opt-multihoming-00
> > 
> > 
> > I am saying that I want to you be aware that shim6 is 
> approaching the 
> > problem with a particular framework (shim6) in mind.
> > 
> > if your contribution is relevant to that framework then of 
> course this 
> > is good - if you head off into alternative multi-6 
> solutions that have 
> > nothing to do with shim6, then maybe you should reconsider 
> whether to 
> > cc shim6 in your notes.
> > 
> >    Geoff
> > 
> > 
> > At 10:46 AM 9/06/2006, CTO WEN Haibo wrote:
> > >Hi Geoff,
> > >
> > >I am wondering the exact meaning of your email. Are you 
> meaning that 
> > >some discussions in this particular thread have less relevance with
> > >shim6 or the whole thread has less relelevance to shim6?
> > >
> > >This document tries to solve the source address selection and exit 
> > >router selection, which is related to the following 
> working item from 
> > >the charter of shim:
> > >"In addition to the network layer shim solution, the shim6 WG is 
> > >specifically chartered to work on:
> > >  Solutions for site exit router selection that work when each ISP 
> > >uses ingress filtering, i.e. when the chosen site exit needs to be 
> > >related to the source address chosen by the host. This site exit 
> > >router selection and the associated address selection 
> process should 
> > >work whether or not the peer site supports the shim6 protocol."
> > >
> > >Best regards,
> > >
> > >Haibo
> > >
> > > > -----Original Message-----
> > > > From: Geoff Huston [mailto:gih@apnic.net]
> > > > Sent: 2006å¹´6æo^9æ-¥ 07:00
> > > > To: CTO WEN Haibo; Greg Daley
> > > > Cc: Lawrence Zou; shim6@psg.com
> > > > Subject: RE: RE: questions about
> > > > draft-wen-ipv6-rsra-opt-multihoming-00
> > > >
> > > >
> > > > Could please I remind all of you in this particular mail thread 
> > > > that the charter of shim6 is to refine and develop one 
> particular
> > approach to
> > > > multi-homing, and it is definitely NOT to rehash over the
> > > > multi6 discussion
> > > > of "all possible approaches to multihoming".
> > > >
> > > > Accordingly, please think for a second about the 
> relevance of your 
> > > > contribution to shim6 before including shim6 in your cc list.
> > > >
> > > > thanks,
> > > >
> > > >    Geoff
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > At 03:38 PM 7/06/2006, CTO WEN Haibo wrote:
> > > > >Hi Greg,
> > > > >
> > > > >Please see the response below. Thanks.
> > > > >
> > > > >Haibo
> > > > >
> > > > > > -----Original Message-----
> > > > > > From: Greg Daley [mailto:Greg.Daley@eng.monash.edu.au]
> > > > > > Sent: 2006?ê6??7è? 11:15
> > > > > > To: CTO WEN Haibo
> > > > > > Cc: Lawrence Zou; multi6@ops.ietf.org; shim6@psg.com
> > > > > > Subject: Re: RE: questions about  
> > > > > >draft-wen-ipv6-rsra-opt-multihoming-00
> > > > > >....
> > > > > > > The core of this multi-homing information option is
> > > > that to provide
> > > > > > > someextra information along with prefix information
> > > > option to help
> > > > > > > host do
> > > > > > > selection, and the host uses the prefix once it finds
> > > > the matching
> > > > > > > information in multi-homing information option.
> > > > > > > ISP name is a sub-option, ISP name is a generalized
> > > > name, the IPTV
> > > > > > > service provider can name itself as "IPTV", Internet
> > > > > > service provider
> > > > > > > can name itself as "HSI", these names also indicates
> > > > their service
> > > > > > > type.Maybe I haven't made this very clear in the draft.
> > > > Of course,
> > > > > > > any other option can be also defined under this
> > > > architecture, for
> > > > > > > example, what
> > > > > > > kind of service the isp can provide, the qos 
> parameter, etc.
> > > > > > > Currently,only a few suboption have been defined.
> > > > > > >
> > > > > > > Using this multi-homing option, router can priodically
> > > > advertises
> > > > > > > the
> > > > > > > ISP's ability and related information.
> > > > > > >
> > > > > >
> > > > > > I think it's a much better fit to provide this
> > > > (changeable) operator
> > > > > > related information using 802.21 media independent 
> information 
> > > > > > services.
> > > > >802.21 is developing standards to enable handover and
> > > > interoperability
> > > > >between heterogeneous network types including both 802
> > and non 802
> > > > >networks. 802.21 is running below IP layer. However, SAAC
> > > > procedure is
> > > > >running in IP layer.
> > > > >You know, in multi-homing site, each router will advertise
> > > > RA which contains
> > > > >IPv6 prefix information, and the interface of host will
> > > > receive all this
> > > > >RA. The
> > > > >standard Stateless Address Auto-Configuration doesn't
> > provide extra
> > > > >information to help prefix selection. Multi-homing
> > > > informaiton option is for
> > > > >this purpose.
> > > > >
> > > > > >
> > > > > > With respect to access network related configuration
> > information,
> > > > > > we already have DHCPv6 for that, and more specific
> > route options
> > > > > > for Router Discovery.
> > > > >In IPv6, there are two ways for address auto-configuration,
> > > > i.e., stateful
> > > > >configuration based on DHCPv6, and stateless configuration.
> > > > It seems there is
> > > > >no standard that says only DHCPv6 can be used in access
> > > > network and SAAC
> > > > >is forbidden.
> > > > >
> > > > > >
> > > > > > Router Discovery is for prefix and next hop
> > > > determination.  Let's not
> > > > > > make it complicated for other reasons.
> > > > >More information along with prefix information option in RA,
> > > > will help
> > > > >host select
> > > > >the appropriate prefix to form its IPv6 global address and
> > > > also select its
> > > > >exit
> > > > >router. It's not trying to make things complicated. Without
> > > > this effort,
> > > > >host cannot
> > > > >find the appropriate prefix and exit router.
> > > >
> > > >
> > > >
> > 
> > 
> > 
> 
>