Re: [addr-select-dt] slide to 6man presentation

[Arifumi Matsumoto <[e-mail]>]

On 2010/07/26, at 21:35, Tony Hain wrote:

> I can present it, and don't have any further comments other than a TBD
> reference for pulling the 'try addr sel again' ICMP from
> draft-hain-ipv6-fwrh-03.txt

I don't object to add a reference to that.
But, as the design team, we have not examined what that approach can solve and cannot.


And, personally I have a working group item titled draft-ietf-6man-addr-select-sol that covers possible approaches to the address selection problems. One of which approach is very similar to ICMP error based approach.

Though I cannot argue about what is not yet made public, if it is similar, that approach does not cover all the cases that were raised so far, i.e. destination address selection related cases. And several other issues it has also, like that error-message-driven approach forces a user/host to re-connect to the same host.

So, as a method to update policy in host, IMO, distributing policy-table approach should be the way forward.

>> -----Original Message-----
>> From: [e-mail] [mailto:addr-select-dt-
>> [e-mail]] On Behalf Of Arifumi Matsumoto
>> Sent: Monday, July 26, 2010 11:31 AM
>> To: Tim Chown
>> Cc: [e-mail]
>> Subject: Re: [addr-select-dt] slide to 6man presentation
>> 
>> Thank you for comments.
>> I'll update.
>> 
>> BTW, I saw Tony here in Maastricht.
>> Can you present ??
>> 
>> 
>> On 2010/07/26, at 18:13, Tim Chown wrote:
>> 
>>> Hi,
>>> 
>>> Some comments....
>>> 
>>> - the DT draft is missed from the slides: draft-ietf-6man-addr-
>> select-considerations-02
>>> 
>>> May wish to have a slide summarising the most recent focuses in that
>> draft, for example including some/all of:
>>> 
>>> - frequent policy changes are linked to having either rapid traffic
>> engineering changes/external routing changes or lots of mobile nodes
>> within the site moving between points of attachment with different
>> policies
>>> 
>>> - it seems sensible to acquire RFC3484 policy when acquiring other
>> configuration information
>>> 
>>> - in a managed enterprise there is likely to be a managed policy, and
>> DHCP available
>>> 
>>> - in dynamic/unmanaged networks using routing hints for address
>> selection *may* be more appropriate
>>> 
>>> - the problem of handling policy conflict is a host issue, the method
>> to distribute the policy is a network issue - we focus here on the
>> network issue since the host issue is broadly applicable to many
>> configuration parameters
>>> 
>>> - we should avoid delaying progression of a 3484 policy distribution
>> method applicable to managed enterprise networks
>>> 
>>> I think this material could fit in a new Slide 3 and then you
>> continue the arifumi-6man-addr-select-conflict reference after that, to
>> keep the flow.
>>> 
>>> I see a new Troan multihoming draft came out this morning, I've not
>> checked what changes there are yet.
>>> 
>>> Cheers,
>>> Tim
>>> 
>>> On 24 Jul 2010, at 23:49, Arifumi Matsumoto wrote:
>>> 
>>>> Hi,
>>>> 
>>>> I drafted the slide for 6man.
>>>> This version still has some clear errors, so
>>>> I have to fix them.
>>>> 
>>>> I really want you all to comment on this.
>>>> About the story itself, wording, typos, and whatever.
>>>> 
>>>> <ietf78-6man-dt.pdf>_______________________________________________
>>>> addr-select-dt mailing list
>>>> [e-mail]
>>>> https://www.ietf.org/mailman/listinfo/addr-select-dt
>>> 
>>> _______________________________________________
>>> addr-select-dt mailing list
>>> [e-mail]
>>> https://www.ietf.org/mailman/listinfo/addr-select-dt
>> 
>> _______________________________________________
>> addr-select-dt mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/addr-select-dt
>