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

RE: New (-02) version of IPv6 CPE Router draft is available for review



Alain,

Please go back to this draft's emails and see how Wes, I, and Remi
discussed the ULA case and a MUST requirement.  Your idea below is not
valid because you have not considered the case of a standalone CPE
Router that is purchased from retail.  The user will power on the CPE
Router at home and the LAN interface(s) of the router will be up first -
the user hasn't connected the WAN port yet to the broadband network.
The user needs some IPv6 address to configure the CPE Router via the
web.  That is why the ULA will be used.  We have also said in our draft
that manual configuration is out of scope of our draft - therefore we
are not worried about details like how does the user get to know the ULA
prior to logging on to the CPE Router over the web using this ULA.  If
need be, we can also work on such details but there is no dispensing
with ULA.  I too like Brian's suggestion of once per factory-reset and
further I prefer a well-known ULA address publishes apriori in the CPE
Router manuals for the user to use that address to configure the CPE
Router.  After configuration, any other ULA may be generated and the
router reset to use the new ULA and associated ULA prefix(es).

Only for the case when the CPE Router is embedded in a broadband device
like a modem and thus managed by the SP, the ULA's will not be spawned
as this is a WAN init first before LAN init model of our draft.  See my
other comments about WAN init first model also complying with CableLab
docsis 3.0 standards. 

Hemant  

-----Original Message-----
From: Alain Durand [mailto:alain_durand@cable.comcast.com] 
Sent: Friday, July 18, 2008 11:53 PM
To: Brian E Carpenter
Cc: Hemant Singh (shemant); v6ops@ops.ietf.org; Wes Beebee (wbeebee)
Subject: Re: New (-02) version of IPv6 CPE Router draft is available for
review




On 7/18/08 11:42 PM, "Brian E Carpenter" <brian.e.carpenter@gmail.com>
wrote:

>> I would rather like the text to recommend to only use ULA when 
>> nothing else is available and immediately renumber to GUA when those
are acquired.
> 
> IMHO that is only acceptable if there MUST be a configurable option to

> support a ULA (generated once per factory-reset) in parallel with a 
> GUA, if that's what the user wants. I think that is needed for 
> scenarios where people want addresses for devices that are guaranteed 
> never to reply to external packets.

I'd agree to this if, by default, parallel ULA support is off.

  - Alain.