RE: draft-wbeebee-ipv6-cpe-router-04 comments

["Hemant Singh (shemant)" <shemant@cisco.com>]

In line below between <hs-wb> and </hs-wb>.

Section 5 "Basic IPv6 Provisioning" says:
> The CPE Router MUST support at least one of two WAN interface  
> models, one of which will be active on the CPE Router at any given  
> time.
p1. If the CPE router only supports one of the two, then only one will  
be active at any given time.

p2. If I were running an ISP, I'd be very suspicious of this  
requirement on CPE router vendors that they MUST support "at least one  
of two" and not both WAN interface models.  Fortunately, I'm on the  
other side of that fence, so I'm very pleased to see that I can refuse  
to support the Unnumbered Model and still comply with Best Current  
Practice.

<hs-wb>
This is exactly what the CableLabs specified eRouter does.
</hs-wb>

p3. I'm deeply confused about how the default route is communicated to  
the CPE router over its WAN link from the service provider in the  
Unnumbered Model.

<hs-wb>
Numbered or Unnumbered models aside, the CPE Rtr receives an RA from the
neighboring first hop SP router. CPE Rtr receives the RA and knows this
is the IPv6 default router.
</hs-wb>


-----
Section 5.4 "Process RAs"

Do these include RFC 4191 More Specific Route advertisements?  Yay!   
That will make the rogue RA problem a whole lot more fun.


<hs-wb>
This section clearly discusses only the direction where an RA comes to
the CPE Rtr from the SP router to the home.  For a cable and DSL network
the SP to the cable modem or DSL modem link is cable RF transport or DSL
telco network.  Such networks are deemed very secure.  A Rogue RA is not
expected in this direction.   Outside of this section of our document,
we can totally agree that a rouge RA may be sent from say, a Windows or
MAC personal computer attached to the CPE Rtr in a direction upstream
from the CPE Rtr to the SP network.  A CPE Rtr has to block any RA from
the home to the SP network.  
</hs-wb>
-----


I'm still going through the draft.  Sorry for my slow progress.  Sigh.

<hs-wb>
No worries.  Comments are welcome any time.
</hs-wb>

Hemant & Wes



--
james woodyatt <jhw@apple.com>
member of technical staff, communications engineering