Re: draft-van-beijnum-multi-mtu-05.txt

[David Lamparter <[e-mail]>]

On Wed, Apr 06, 2016 at 01:12:45PM -0400, Michael Richardson wrote:
> David Lamparter <[e-mail]> wrote:
>     > Now the hard bit; L2 devices messing with L2 addresses.  The device
>     > class I have in mind here are 802.11 "client to wired" bridges.  They
>     > connect to wifi as plain clients and have one or more LAN ports
>     > supporting multiple devices.  802.11 restricts them to the single MAC
>     > address their radio has, so what they do is they "MAC-NAT" everything
>     > behind them.  (They rewrite ARP & ND and have an address table to undo
>     > it on incoming wireless packets.)
> 
> This definitely falls into the category of stupid-layer-2 tricks to get
> around a lack of a routed (homenet-style) layer-3.  sigh.  IPv4 legacy.

Indeed, though it's also 802.11 failing to cover bridges as plain
clients.  802.11ak (= work in progress) and non-standardised WDS (widely
available) solve the same issue, but they need AP implementation (11ak)
or even configured cooperation (WDS); thus that class of kludge products
will stay around for quite a while...

> It seems you are saying with can deal with this problem with:

Yes (sorry if I worded this poorly) - I arrived at the suggestion below
in the hope that identifying by link-locals is less prone to L2
shenanigans than identifying by link-layer address.

> > I believe the better choice here is "Neighbors are identified by their IPv6
> > link-local address".  This removes any possibility for 2 hosts to be
> > conflated into a single neighbor even in the presence of 'broken' L2
> > devices.