Slashdotters defend IPv6 from UK blogger who still thinks its no big deal(tech.slashdot.org)
tech.slashdot.org
Slashdotters defend IPv6 from UK blogger who still thinks its no big deal
http://tech.slashdot.org/story/10/08/18/1610230/Why-You-Shouldnt-Worry-About-IPv6-Just-Yet
3 comments
It's not necessarily a binary thing, which I think is one of the problems people have when discussing this. We'll never actually run out of IPv4 addresses. There's 4 billion-ish of them and they aren't disappearing into the ether or anything. It's just another limited resource the market can deal with, and for the right price you'll always be able to buy a block of IPv4 addresses.
Mind you, that "right price" will start going up pretty quickly. Supply is fixed, but demand is growing pretty quickly. To some extent you can't see how fast the demand is actually rising because it's already hidden behind large NAT installations and such. The usual microeconomic equations for computing this sort of thing usually have a built-in assumption that for the right price production can be increased; when production absolutely can't be increased the price tends to rise faster than you'd expect. I don't know that there are very many blocks that are allocated but not actually used anymore, you'll be paying people to drop their block and pay the price of transitioning to a NAT setup or something, which can be quite painful.
So people watching the free-IP-block count aren't really watching the right number. The number to watch is the price of an IP block. Since I think the price of an IP block right has discounted into it the fact that we're moving to IPv6 and the crisis will be averted, it probably isn't as high as it would be if we weren't expecting to move to IPv6.
And so the real problem isn't really the "running out". The real problem is that IPv4 address as a limited, and therefore expensive, resource is just stupid. Yes, the market could allocate it efficiently, but it's better for everybody if it simply isn't effectively limited in the first place.
And then the consequence of that is that you end up with a lot of people talking past each other; "well, if it isn't running out what's the problem then?" The problem is that in 2016 a startup's largest expense shouldn't be their public IP block.
Mind you, that "right price" will start going up pretty quickly. Supply is fixed, but demand is growing pretty quickly. To some extent you can't see how fast the demand is actually rising because it's already hidden behind large NAT installations and such. The usual microeconomic equations for computing this sort of thing usually have a built-in assumption that for the right price production can be increased; when production absolutely can't be increased the price tends to rise faster than you'd expect. I don't know that there are very many blocks that are allocated but not actually used anymore, you'll be paying people to drop their block and pay the price of transitioning to a NAT setup or something, which can be quite painful.
So people watching the free-IP-block count aren't really watching the right number. The number to watch is the price of an IP block. Since I think the price of an IP block right has discounted into it the fact that we're moving to IPv6 and the crisis will be averted, it probably isn't as high as it would be if we weren't expecting to move to IPv6.
And so the real problem isn't really the "running out". The real problem is that IPv4 address as a limited, and therefore expensive, resource is just stupid. Yes, the market could allocate it efficiently, but it's better for everybody if it simply isn't effectively limited in the first place.
And then the consequence of that is that you end up with a lot of people talking past each other; "well, if it isn't running out what's the problem then?" The problem is that in 2016 a startup's largest expense shouldn't be their public IP block.
If I want an IP block (and have a legit reason), I can get a previously unallocated block. Soon, there are going to be no unallocated blocks. We will have _run out_ of unallocated blocks.
(That we are able to re-allocate, and thus, resell previously allocated blocks/addresses only means that there is a market, after all blocks have been allocated.)
(That we are able to re-allocate, and thus, resell previously allocated blocks/addresses only means that there is a market, after all blocks have been allocated.)
The issue as well, not just with companies having to NAT now, is that some of them are using their IP block for internal IP addresses that are already NAT'ed. Changing entire network infrastructures over to non-public routable is a costly expense as well.
Hurricane Electric have an IPv4 exhaustion JS based countdown widget on the front page of http://www.tunnelbroker.net/
On May 28th of this year I wrote an article:
https://secure.grepular.com/Setting_up_IPv6_is_Easier_Than_Y...
At the time of writing, they were predicting we had 427 days left. Today they're predicting we have 281 days left. It's gonna bite a lot of people on the ass.
On May 28th of this year I wrote an article:
https://secure.grepular.com/Setting_up_IPv6_is_Easier_Than_Y...
At the time of writing, they were predicting we had 427 days left. Today they're predicting we have 281 days left. It's gonna bite a lot of people on the ass.
This was meant to be a reply to the comment by tomjen3.
My bad. :)
My bad. :)
Is the bloggers location in any way relevant to the article?
It's relevant because Europe is the standard example of "we need ipv6". Americans largely don't care. :)
Yeah, I thought mentioning that he was from the UK is a bit strange too.
While I don't see any point any point in turning ipv6 of, I also don't think we are going to run out of addresses anytime soon - it would be nice if we could end natting (It would certainly solve a few of my problems), but just like with the oil situation, the doom sayers are, as usual, way out there.
For a reference, we have been running out of oil since the early 1910.
For a reference, we have been running out of oil since the early 1910.
We're due to run out of it within a year. We know how many blocks remain, and how fast they are allocated. It's not that hard to tell when we're going to hit zero...
The amount of oil in the world is speculated on. The number of IPv4 addresses is numerically fixed and known.
As other commenters have noted, it's not that one day your computer will just inexplicably fail to pull an IP address. But they are suddenly going to be come a more precious (read:expensive) resource.
As other commenters have noted, it's not that one day your computer will just inexplicably fail to pull an IP address. But they are suddenly going to be come a more precious (read:expensive) resource.
[1] http://arstechnica.com/tech-policy/news/2010/07/internet-wil...