Re: [homenet] Let's make in-home ULA presence a MUST !?

[James Woodyatt <[e-mail]>]

On Wed, Oct 22, 2014 at 12:51 PM, Ted Lemon <[e-mail]> wrote:

> On Oct 22, 2014, at 2:46 PM, James Woodyatt <[e-mail]> wrote:
> > They may often be the only *default* routers, but there can be— and
> absolutely definitely will be in the vast majority of cases— overlay
> networks that route ULA prefixes to, from and most likely *between* home
> networks over tunnels. We can't tell people not to do that. If there is a
> routing protocol in a HOMENET, then it will be done, and it ought to work
> right.
>
> In the case where ULAs are being routed like this, wouldn't that ULA be
> the responsibility of whatever homenet federation protocol is being used?
>  I don't disagree that this is a valid use case, but I don't think it would
> rely on the homenet ULA.
>

My point is that it doesn't need to be done that way unless HOMENET forces
that design choice.

I see a way to work around the potential problem here— by eating the
expense of requiring the overlay routers between HOMENET site boundaries to
exchange the full raft of valid /64 routes in all the currently valid
locally-generated ULA prefixes instead of exchanging just the aggregated
/48 ULA prefixes. I suppose that can be made to mostly work in the majority
of cases at the cost of memory and efficiency for interior routers. Don't
grow your home networks too big, however, or the interior routers in your
house— or in your friend's house— might fall over when the overlay connects.


-- 
james woodyatt <[e-mail]>
Nest Labs, Communications Engineering