Sunday, 6 February 2011

IPv4 and 6

Right, last post for today, making it a record 3 in a day. I have a couple of other topics to post on, but seeing as how its 2am, I'm going with the easy one. So what is IPv4? Glad you asked!

If you are reading this, you are connected to the Internet. But what you may not have ever given any thought as to how this is possible. You may know that you plug the LAN cable from the router into your Ethernet port or connect to your wireless network, but what really goes on?

The Internet (capital 'I') came out of the concept of internets (small 'i'), which is a network of networks. So how do you know who is who and who is on which network? Simple assign them all a unique identifier. Then the question arises as to how you do that, because we now have the problem that everybody needs to understand how these identifiers work, so we need a common language, if you will.

And thus was born the Internet Protocol version 4, aka IPv4. Defined in RFC 791, the Internet Protocol is how all devices connected to the Internet identify each other and communicate to each other. So, what's the problem?

IPv4 address are 32-bit address broken into 4 8-bit groups called octates. Now this means there is a finite number of these addresses, approximately 4 billion, which we are now out of. Certain ranges of addresses are restricted for specific purposes, but it is just a small portion. As of February 3rd, all IPv4 addresses had been assigned by the Internet Assigned Numbers Authority (IANA) to the Regional Internet Registries (RIRs), when they assigned the last 5 remaining /8 blocks, which is a set of addresses with the first octate fixed.

This is a real problem, as at some point, new devices will not be able to connect to the Internet. There does exist a solution: IPv6, as defined in RFC 2460 & RFC 2373. Version 6 addresses are 128-bits, broken into 8 hexates or 16-bit groups, as compared to the 32-bit version 4 addresses. This gives us a hugely greater number of addresses and would solve this problem of address exhaustion. Well, there is a small catch.

Despite IPv4 essentially having being exhausted, IPv6 is still not implemented fully. So we are currently in a weird transition period where things are a bit muddled. Almost everybody has implemented support for both IPv4 and IPv6, but there is no strict adherence to IPv6.

My issue with this is that the existence of a parallel legacy system has almost always created some kind of security threat that the new system can not deal with. To my knowledge, there is now such security loophole in IP, yet. I'm sure somebody, somewhere will find something and start exploiting it. It may not be a big hole, but it will probably be there. I would truly be happy to be proven wrong and hopefully we will transition over to IPv6 without incident.

On that note, that is all from me today. I will try and get as many of the latest stories, but they seem to be cropping up faster than I can handle.

*******EDIT*********
So, I may have said I will be catching up on stuff, which I totally am! But, also this came to my attention.

No comments:

Post a Comment