On Jul 25, 2013, at 10:20 AM, Nick Hilliard <nick@inex.ie> wrote:
that is true, but it might be useful to make some sort of estimate particularly with regard to the class A+B legacy allocations. Eyeballing, I figure there's probably quite a good chunk of unrouted legacy space or legacy space which ended up in commercial organisations who probably don't need all that space, and which could eventually come into the market - maybe 30-40 /8s as a rough estimate. Given that the RIRs ploughed through 94 /8s since the early 1990s, this represents a reasonably large amount of integers from the point of view of potential future liquidity.
Nick - It's probably best to look at the last 2 or 3 years of IANA->RIR allocations as an estimator of demand, rather than averaging across the entire period, e.g. reference Geoff Huston's recent blog entry for some very interesting pre-runout statistics (including APNIC numbers in excess of 12 /8's per year) - http://www.potaroo.net/ispcol/2013-08/when.html In terms of available space that may come to market, I'll note that many of these address blocks are actually in use by organizations, but in terms of routing hidden behind various firewall/nat combinations. Taking both the recent burn rate and in-use-but-not-routed factors into consideration, it's likely that (at a high enough price) that there are several years of additional IPv4 liquidity that could be obtained, but rather uncertain as to whether it could provide, say, 10 more years of continued IPv4 Internet growth... FYI, /John