No new IP adresses by next year? | Hacker News Reader

No new IP adresses by next year?

telegraph.co.uk

18 pointsmikefrancis11 comments

It's a bit funny observing those predictions. Everyone has his own... Using a not very scientific probing on google, This is the number of articles saying that we will run out of ipv4 in year X:

- 2007 - 689

- 2008 - 1,970

- 2009 - 618

- 2010 - 3,590

- 2011 - 2,180

- 2012 - 1,750

- 2013 - 2,620

- 2014 - 749

- 2015 - 686

I wonder if there are any betting companies interested in this event ;)

You can meet all your IPv4 exhaustion counter needs here: http://ipv6.he.net/statistics/

This is also a good source if you want to know how the exhaustion will occur: http://ipv4.potaroo.net/

Good. Companies are going to keep putting off IPV6 migrations until ARIN slows down their deployments of new IPV4 IP space. To this point all the doomsday talk hasn't had much impact on reality. Request IPs, get IPs, ignore problem. Once ARIN does slow down new deployments companies will have to make more efficient use of their existing IPV4 space in the short term. Most of the big deployments I've seen in ISP environments are very wasteful or were hugely over engineered to start with. In my environment I haven't requested new IP space in the last 3 years and could probably go another 2-3 years without a single new public IPV4 IP address.

I think they know what they're doing by requesting more... At least in Europe you have to properly label your IPs and that way "prove" they're used. You will not get new ranges just because you request them.

Solution? Every time a new service goes live - give it a new ip. And another one for failover. You fill your allocations quickly and go back for more ips. This way people who over-allocated will have no problem to free up a full /24 when they need it, once ARIN and others slow down. But you won't get a new /24 at all and will not have any space to reuse / redesign... Cruel, but that will happen.

>> "We'll be down to our last tens of thousands or hundreds of thousands of web addresses by the end of next year,"

I don't buy it. There's millions of unused IP addresses, mostly massive blocks that were allocated to big companies early on. If IPv4 addresses truly become scarce, these big companies will simply start selling some off.

Not saying it's a bad idea to switch over, but it's not something most people need worry about for a few years IMHO.

We've been hearing this for years now. It's like when I was at school they told us we'd run out of oil in 2000 or something.

Large chunks of the large early allocations have already been handed back over the last few years. There isn't that much more left to reclaim or sell.

You cannot sell them, at least openly. If they cached you, you could lost all your allocations.

> If they cached you, you could lost all your allocations.

Is this because of a poor serialisation format?

Unfortunately IPv6 was built without proper backwards compatibility. They should have simply made it so that every IPv4 address automatically was an IPv6 address, and set up IPv6 -> IPv4 NAT so that entire IPv6 networks could easily hide behind a single IPv4 address. But neither was initially done.

Without these two obvious measures, nobody wants an IPv6 address because it is useless to have one. I have no idea why this wasn't obvious to the IETF, but apparently they missed it.

At least they now admit that this was a mistake: http://www.networkworld.com/news/2009/032509-ipv6-mistake.ht...

Oh and the whole "no NAT" thing? I'm not looking forward to that at all. You see the top reason for using NAT in homes is so that your internet provider can't charge you for putting multiple lightly used computers on your connection. And they aren't trying to charge you because they are running out of IP addresses, they are doing it because it is a way to identify high-end customers who can be charged more. And IPv6 will close a big loophole for them.

Damn. I just locked up 2 of 'em.

Reply on news.ycombinator.com