On Saturday, September 17, 2011 06:21:07 PM Usman Latif
wrote:
To me it just seems like a crazy idea of assigning a /64
subnet (that can otherwise fulfill requirements of
18,446,744,073,709,551,616 hosts) to a residential CPE
which today uses at the max 5-10 IPv4 addresses on their
LANs.
It is, obviously, wise to remember that even as existing v4
customers deploy v6, the size of their networks isn't going
to necessarily grow ten-fold by them doing so.
In the field, we've seen customers concerned about not
deploying v6 because they cannot find a decent firewall
which will support v6 NAT (that's right, NAT66 - who'da
thought with v6 aplenty?) without them having to test for
months or years. Yes, there are solutions out there already,
but support is still spotty when compared to v4.
So if users aren't going to suddenly be assigning public v6
addresses to their printers anytime soon, and they won't be
rampantly scaling up network devices just because they
suddenly got a /56 or /48, it's interesting that we expect
them to jump from using 1000 v4 addresses, for example, to
billions with v6, in one step.
But that argument is probably too simplistic :-).
I think because we have been working with IPv4 and have
been really careful about not wasting IPs due to the
limited address space, maybe its that same mindset which
makes me uncomfortable looking at multiple /64s going to
a single DSL end-customer etc. :)
On the other hand, perhaps all the suffering we have gone
through with v4 should tell us something about how to
"spend" v6.
v4 projections, during the "experiment", were based on
trends of the day. They simply couldn't have predicted that
32 bits in v4 wouldn't be enough a few decades later.
v6 projections today, for the most part, are based on
"current" trends in most networks. Who knows what will
happen 2, 3 or 4 decades from now that could simply blow our
trends of today out the water?
But again, I suppose this argument is too simplistic :-).
Cheers,
Mark.