diff options
author | Nicholas Johnson <nick@nicksphere.ch> | 2022-05-23 00:00:00 +0000 |
---|---|---|
committer | Nicholas Johnson <nick@nicksphere.ch> | 2022-05-23 00:00:00 +0000 |
commit | 05fa3051e12acddfe320912a93e1927bcf1b64f6df2a14589594144df3b9f3e2 (patch) | |
tree | e2f767706bbef2caf24a3fd5ea9147f6866d3fef2c0e732f9b481932e87d67ea /content/entry/ipv6-adoption.md | |
parent | 44ef9882132619ead1f888778804893d848b7686a4833e038b67b263165eb933 (diff) | |
download | journal-05fa3051e12acddfe320912a93e1927bcf1b64f6df2a14589594144df3b9f3e2.tar.gz journal-05fa3051e12acddfe320912a93e1927bcf1b64f6df2a14589594144df3b9f3e2.zip |
Fix spelling errors
Diffstat (limited to 'content/entry/ipv6-adoption.md')
-rw-r--r-- | content/entry/ipv6-adoption.md | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/content/entry/ipv6-adoption.md b/content/entry/ipv6-adoption.md index 85125cb..268f30a 100644 --- a/content/entry/ipv6-adoption.md +++ b/content/entry/ipv6-adoption.md @@ -14,12 +14,12 @@ This was fine when the internet was small, but now the internet is massive and h Welcome to NAT[2]. NAT stands for Network Address Translation. The main reason NAT exists is to solve the IPv4 problem of not having enough logical addresses for every device. NAT translates private IP addresses on an internal network to public IP addresses that can talk to other computers on the real internet. This allows several connected devices to share the same IP address, conserving logical addresses so IPv4 can still work. I won't go into detail on how this happens because it's not relevant, but it does have overhead. NAT is basically an ugly hack for the problem of not enough IPv4 addresses for each internet connected device. # IPv6 -IPv6[3] supercedes IPv4 using 128-bit addresses (340 undecillion IP addresses). It's the obvious elegant solution to the problem of not having enough internet addresses: use a protocol that has more addresses. It doesn't require NAT because each connected device can have its own IP address on the real public internet. Since the IPv6 address space is so huge, it's highly unlikely that IPv6 will ever be superceded for lack of internet addresses. +IPv6[3] supersedes IPv4 using 128-bit addresses (340 undecillion IP addresses). It's the obvious elegant solution to the problem of not having enough internet addresses: use a protocol that has more addresses. It doesn't require NAT because each connected device can have its own IP address on the real public internet. Since the IPv6 address space is so huge, it's highly unlikely that IPv6 will ever be superseded for lack of internet addresses. It also has other practical advantages to IPv4. As the name implies, it's a newer protocol drafted in 1998 whereas IPv4 was first deployed in 1982. IPv6 packets are easier for routers to process since the IPv6 packet is simpler than the IPv4 packet. This is consistent with the original vision of the internet where most processing happens at endpoints, not routers. IPsec[4] is mandatory whereas in IPv4 it was retrofitted. Network operators don't have to do port forwarding on the router or make firewall changes. Multicast addressing is simpler. IPv6 limits the size of routing tables[5]. Mobile IPv6[6] is as efficient as regular IPv6. I could go on but the point is it's much better than IPv4 in every way. # IPv6 Adoption -ISPs and tech giants are slowly increasing IPv6 support. Ideally, everyone would use IPv6 and IPv4 would cease to exist. IPv4 has no practical advantages. It was superceded by IPv6 over 2 decades ago and the switch still hasn't completely happened yet. What's the problem? If IPv6 is better then why is adoption taking so long? The barrier to IPv6 adoption isn't so much at endpoints. By 2011 all major operating systems had support for IPv6. The problem is there often isn't a strong financial incentive for IPv6 adoption. +ISPs and tech giants are slowly increasing IPv6 support. Ideally, everyone would use IPv6 and IPv4 would cease to exist. IPv4 has no practical advantages. It was superseded by IPv6 over 2 decades ago and the switch still hasn't completely happened yet. What's the problem? If IPv6 is better then why is adoption taking so long? The barrier to IPv6 adoption isn't so much at endpoints. By 2011 all major operating systems had support for IPv6. The problem is there often isn't a strong financial incentive for IPv6 adoption. If you're an average internet user, you don't even know what IPv4 or IPv6 is. Unless your ISP enabled IPv6 for you then you probably don't have it. You can access all the internet resources you want without it anyway. Even if your ISP enabled it and your modem/router supports it, still many end-user devices and applications don't work well with it. If they do support IPv6, they also support IPv4 because IPv6 always runs alongside IPv4 with dual stack[7]. If you host any internet resource then all your users support IPv4. So why bother with IPv6? |