Switzerland Tops IPv6 Adoption Charts; US Lags At 4th 155
hypnosec writes "According to recent statistics, Switzerland has topped the IPv6 adoption charts by leapfrogging Romania, which led the charts for nearly a year. According to Google, Switzerland's adoption stands at 10.11 percent — the highest for any country. Romania, on the other hand, has an adoption rate of 9.02 percent, followed by France at 5.08 percent. Switzerland took the top position near the end of May and the primary reason seems to be Swisscom and its drive to adopt the next IP version. The U.S. stands at fourth place with just 2.76 percent adoption."
Re:Lags? (Score:5, Interesting)
The linked article seems more than a little odd; I just checked Japan at the same place they link to, and it has an adoption rate of 3.13%, ahead of the US. So it seems the comparison is only among a restricted set of countries (the linked page has only five countries displayed), and not really relevant to much of anything.
Re:Lags? (Score:3, Interesting)
The largest economy in the world is the EU, not the US.
And I'm pretty tired of this argument that its okay for the US to be lagging in so many things because they are big. The US has gobs of resources and a very high GDP per capita. As someone previously pointed out, why can't you find a city or small state with higher IPv6 adoption than Switzerland? It's not like New York is somehow being held back because Los Angeles exists.
It's like that same tired argument that size is why bullet trains are impossible in the US. The east coast is just as densely populated as Europe, if not more, yet there is no decent high speed link between Boston/NYC/Philadelphia/Washington. That's shorter than the French TGV line. People hate the TSA, the population density is there to support high speed rail, but the political will is not.
PS: I think the real reason IPv6 adoption in the US is low is simply because they have so the lion's share of IPv4 blocks already...
Re:Another "crisis" that isn't. (Score:4, Interesting)
That part is true. But back then allocations only came in three sizes. Those allocations really were of the smallest size, which would cover their need. That practice was changed soon enough to avoid problems. Slowing down the allocation of IP addresses and not having any of the already allocated addresses handed back would have given enough time, that IPv6 could have been deployed.
The only problem was, that nobody did. People just kept going on deploying more and more IPv4 networks and ignoring IPv6. Other workarounds came along, which stretched the supply of IPv4 addresses even further. The truth is, those workarounds have caused more problems than they solved. They were not necessary in the first place, there was plenty of time to deploy IPv6. The workarounds mean that we now have a much bigger Internet that needs to be converted, which means more work, and it is more expensive. But worse than that, the workarounds are actually part of the reason transitioning to IPv6 is so damn hard. Had IPv4 been free from any NAT, it would have been easier to have IPv4 and IPv6 co-exist.
Some people suggest those early players should hand back those addresses. It wouldn't solve any problem. It would have delayed the problem by a few months. But the problem would have returned and been just a tad worse. Also, it is a myth that those addresses are unused. Even if they are not all advertised in BGP, they may be used internally on systems, which also need to communicate with the public Internet. Hence they cannot be reused without breaking some communication. And even if they could be handed back, the amount of work it would take to ensure they are really not used plus the administrative overhead, means it is just not worth the effort. All that effort would be better spent working on a real solution.
All of those addresses, which could possibly have been handed back would have been used already in 2011. IANA ran out of addresses in early 2011, and APNIC was growing fast at the time.
That's only true, because they already have. Rationing of IPv4 addresses is happening already, and it is affecting end users. The problems end users experience will get worse over time. But very few people understand the connection between the problems they are experiencing and shortage of IP addresses.
But that won't help. There aren't addresses to reallocate. Extrapolate the curve from before IPv4 addresses and ignore the limit. Then you'll find consumption would reach 200% before the end of this decade. No redistribution of IP addresses will solve that. Also redistribution of IP addresses is a problem in itself. Every time you break up a block and redistribute the addresses, the address space gets more fragmented. This fragmentation means more routing table entries, which consume costly CAM resources on the backbone routers. This is a side effect of stretching the utilization of addresses too far.
Research has shown that you should not expect to utilize more than 80-90% of the bits in an address, if that address is supposed to be used for routing. That means you should not expect to utilize 32 bits of the IPv4 addresses, but only 26-29 bits.