Geoff Huston, chief research engineer at APNIC Internet Registrar, has predicted that IPv4 addresses will run out in 2020. In the new series of materials, we will refresh the information on how the addresses were depleted, who still had them and why it happened.
/Unsplash/
Why do addresses run out
Before moving on to the story of how the IPv4 pool dried up, let's talk a little about the reasons. In 1983, the introduction of TCP/IP used 32-bit addressing. While
At the same time, in the 80s, many organizations received more addresses than they really needed. A number of companies still use public addresses for servers that operate exclusively on local networks. The spread of mobile technologies, the Internet of things and virtualization added fuel to the fire. Miscalculations in estimating the number of hosts in the WAN and inefficient address allocation have caused the IPv4 shortage.
How did the addresses end?
Early XNUMXs APNIC director Paul Wilson
2011 year: As Wilson predicted, the APNIC Internet registrar (responsible for the Asia-Pacific region) has the last
2012 year: The depletion of the pool was announced by the European Internet registrar RIPE. It also started allocating the last /8 block. The organization followed APNIC's example and introduced strict restrictions on the distribution of IPv4. In 2015, RIPE had only 16 million free addresses. Today, that number has dwindled significantly.
A couple of fresh materials from our blog on Habré:
2013 year: Jeff Huston of APNIC on the blog
2015 year: ARIN
2017 year: About stopping the issuance of addresses
2019 year: Today, all registrars have a relatively small number of addresses left. Pools are kept afloat due to the fact that unused addresses are periodically returned to circulation. For example, at MIT
What's next
It is believed that IPv4 addresses
Network Address Translation (NAT) allows you to translate multiple local addresses into one external address. The maximum number of ports is 65. Theoretically, the same number of local addresses can be mapped to one public address (if you do not take into account some limitations of individual NAT implementations).
/Unsplash/
ISPs can turn to specialized solutions - Carrier Grade NAT. They allow you to centrally manage local and external addresses of subscribers and limit the number of TCP and UDP ports available to clients. Thus, ports between users are distributed more efficiently, plus there is protection against DDoS attacks.
Among the disadvantages of NAT, potential problems with firewalls can be identified. All user sessions go online from one white address. It turns out that only one client at a time can work with sites that provide access to services via IP. Moreover, the resource may think that it is under a DoS attack and close access to all clients.
An alternative to NAT is the transition to IPv6. These addresses will last for a long time, plus it has a number of advantages. For example, a built-in IPSec component that encrypts individual data packets.
So far, IPv6
We will talk about this next time.
What we write about in the VAS Experts corporate blog:
Source: habr.com