[LINK] Forum on high speed bandwidth in Australia
kauer at biplane.com.au
Thu May 7 16:21:01 EST 2009
On Thu, 2009-05-07 at 15:45 +1000, Rick Welykochy wrote:
> > > It is believed that the IP4 addresses will run out by April 2010, which will
> > make it necessary for users to move over to IPv6, or the next version that
> > supports a total of 16 billion IP addresses
> What does that last sentence mean? Can IP4 be extended somehow to accommodate
> 16 billion addresses instead of the current 4 billion (approx)?
The sentence means above all that the writer has little clue about IPv4
or IPv6. Taking each in turn:
"It is believed that the IP4 addresses will run out by April 2010"
Current best educated guess is "sometime in 2011", but it is a guess
with huge margins of error. For a complete and exhaustive analysis, see
"which will make it necessary for users to move over to IPv6"
This makes it sound like IPv4 will stop working (not true), that people
will have to stop using it (not true) and move to IPv6 (not true). The
2011 estimate is when the pool of available addresses will run out, no
more. Existing allocations will continue to work fine.
It *is* important that people start looking into and where possible
implementing IPv6 now, or they will soon find themselves unable to
expand, unable to begin new ventures, because there will be no more IPv4
addresses available. Network admins are already feeling the bite and
have been for several years now. IPv4 address scarcity already
complicates mergers and acquisitions. But it is wrong to suggest that
there is a date upon which some sort of Armageddon will commence.
"supports a total of 16 billion IP addresses"
Hm. One single IPv6 subnet supports around 18 billion addresses (64
bits). IPv6 itself supports about 340 trillion trillion trillion
addresses (3.4x10^38). It's hard to put that number into perspective,
but that is more addresses than there are hydrogen atoms in the
And to answer your own question, no. IPv4 cannot be extended to 16
billion addresses. There have been a few suggestions (such as usurping
little-used bits in the IPv4 header), but all of them involve at least
as much disruption as IPv6 does, so not really worth it.
That bit, "IPv6, or the next version" is not saying that there is a next
version other than IPv6, it is saying that IPv6 is the next version.
Very poorly phrased.
Karl Auer (kauer at biplane.com.au) +61-2-64957160 (h)
http://www.biplane.com.au/~kauer/ +61-428-957160 (mob)
GPG fingerprint: 07F3 1DF9 9D45 8BCD 7DD5 00CE 4A44 6A03 F43A 7DEF
More information about the Link