[arin-ppml] IPv6 Allocation Planning

[michael.dillon at bt.com]

> I sat down this morning and tried this.  Its quite difficult to
> overcome
> the ingrained habits that were natural to IPv4.  It will be more
> difficult to convince the bosses to change their thinking.  But we will
> be trying to do so.

You said "bosses". That means that there will be three people at the
meeting. Offer to bring coffee or coke or whatever for that meeting.
Then take your one small coffee, or one can of coke, and three cups
and carefully pour it out, not too fast, and if they say something
to you wave them off and say it's important that you don't spill
a drop because we can't afford to waste beverages.

That's what you are doing when you try to not waste IPv6 addresses.
In fact, there is no shortage of beverages, nor is there a shortage
of IPv6 addresses. This so-called waste, is not waste, it is
investment in painless future network expansion. Every part of
the network has spare addresses so it can expand like an accordion.
If you decide to focus on cloud services in one PoP, then that
PoP already has enough address space to expand into a data center
with 120 virtual machines on every 1U of rack space.

This is why the designers of IPv6 did not just add a few bits to
the 32-bit IPv4 address, they expanded it to 128 bits making it
so outrageously huge that network designers could afford to completely
ignore address conservation and concentrate on simple functional
and expandable network architectures.

--Michael Dillon