RE: Who's afraid of IPv4 address depletion? Apparently no one.

I work at an ISP where i design vpns. The trend the last few years is that vpn`s are moving from crypto to L2, resulting in secure networks with a single egress point.This reduces the amount of needed ips to a bare minimum, and in many cases limited to only DMZ. Nat takes care of userbased internet access

What is this nonsense about 2^128 being a "number is too big for anyone but a mathematician to understand"? Most computing professionals should understand powers of 2... it's what we do! It's so incredibly simple I will explain it here. 2^20 = 1MB, 2^30 = 1GB, 2^40 = 1TB. Notice a pattern? How about this: 2^1 = 2, 2^2 = 4, 2^3 = 8, 2^4 = 16, 2^5 = 32. As you go up one power of 2, the result doubles. So what's it like going from 2^32 (IPv4) to 2^128 (IPv6)? Take the number of IP addresses in IPv4 (2^32, aka about 4.3 billion) and double it... 96 times! (128-32=96) You don't have to be a mathematician to understand this.

Yeah, until you run a website where you're getting 1000s of hits from some IPs, and you can't tell if that's a small business, a hacker at home, or a large enterprise - you will treat the three differently if you're getting that kind of traffic from each. Also, newer IM programs have to use UPnP to try to get a port open (which many corporate firewalls still don't support, even with restrictions).

I've just recently deployeed Apple Airport Extremes that support IPv6 on my network, and we are somewhat cutting edge, so I may turn on IPv6 on those soon.

You sure didn't ask THIS I.T. Professional!

I can't believe you would print such a ridiculous article. It serves just either to inflame or offer more ammo to the idiots who think we can just sit back and keep things as they are.
Back in the day ... we sure worked hard on swapping and paging algorithms, but we sure as heck didn't say "well we don't need large memory cache, that would just be a waste of hardware"!

The big boys on the block are already steaming ahead with IPV6. It has all kinds builtin benefits over IPV4. We'll look back and wonder how we got along on the net with out it. Yeah, there may be even a predominance of folks who just want to put it off, but just like most of the previous net related changes - we will do it, and the net will be the better for it.

The problem with solving all problems with NAT (PAT) is that it enforces the client-server paradigm. The solutions that attempt to work around this restriction will inevitably cost considerably more (TCO) than just making a move towards something that is more scaleable (IPv6).

IPv6 is not a magic panacea, but it is a viable, evolutionary step in the right direction. I realize that when I say peer to peer, most will immediately think about file sharing, but it is more than that. It is a networking methodology. By continuing to constrict ourselves finding additional hacks to extend IPv4, we only serve to limit the possibilities of the next "killer app".

IPv4 address depletion affects the ISP community first, and then larger enterprises who will need to continue growing. This is why 32% and 29% of IT staff in these communities see IPv4 depletion as a "huge concern".

That's actually remarkably high level of awareness for a problem which is still 3 to 4 years out, and it is only going to increase over time. Whether small and medium IT staff need to worry much about IPv4 depletion is predicated on whether their organization is going to be growing and need public address space as an result.

/John