IPv6 Challenges and Opportunities

1sockchuck writes "Opinions differ on when the Internet will run out of IPv4 addresses, prompting a wholesale transition to IPv6. In recent videos, John Curran of ARIN provides an overview of issues involved in the IPv6 transition, while Martin Levy of Hurricane Electric discusses his company's view that early-mover status on IPv6 readiness can be a competitive advantage for service providers. Levy's company has published an IPv4 DeathWatch app for the iPhone to raise awareness of the transition."

Comments

corpspeak to english dictionary

[DaveV1.0]

According to my copy of the CorpSpeak to English dictionary "challenge" and "opportunity" both say "See 'problem'."

marketing speak = teh suck

[Em Emalb]

"Challenges" means problems. "Opportunity" = cool features.

Features of IPv6:

Every known star in our universe can now have 252 ip addresses with ver6.

My frigging socks can tell me they need to be cleaned via a script. My shoes can use GPS to track where I'm going, how many miles I walked/ran that day, etc.

Problems of IPv6: Screw it, we'll just nat our existing IPv4 addresses.

Re:marketing speak = teh suck

[BobMcD]

This, this, o-this-ily-this!

Also I think proponents of IPv6 also tend to overlook the value of DNS. Human short-term memory only has so much space in it. IPv4 addresses tend to be hard to memorize, ergo DNS puts an easy handle on it.

In an IPv6 world you get this memory problem magnified in a huge way:

1) The addresses are now ridiculously long.

2) There's not supposed to be any such thing as NAT (which also means your practice of always having your inside router be x.1 now gets more complex)

3) Many things that don't REALLY need addresses are now going to get them, because we have so many, so lets just go crazy.

To recap, many minor devices will all have a very-long, unique address, and each will be difficult to fit into brain-space alone, let alone together.

This scenario only works in a fully-DHCP world, which is fine for some, but I'll keep my static IPv4 for as long as possible, thanks.

Re:marketing speak = teh suck

[mikael_j]

I'm not sure I'm following you here, so what you're saying is that instead of Joe Q. Sysadmin always having his internal router be 10.0.0.1 and all the hosts having 10.x.x.x IPs tied to hostnames he'd have something like 2001:1001:f00f::1 as the router and all hosts would be in the same subnet? Yeah, that's really scary and confusing...

Also, NAT is an ugly hack that doesn't really need to exist, the packet filtering can be handled with a plain old packet filtering firewall just like it used to be done prior to everyone using NAT and what exactly is the point of address translation? Isn't that like going back to pre-IP days when every network seemed to use its own protocol (or in this case, everyone uses local addresses internally and a single or small number of external addresses) and inter-network communication was a PITA?

And I'd rather see devices that don't need public addresses getting them than "The amazing NAT future" where you have to pay big bucks to get a public IP address instead of being stuck in NAT hell (first they came for the residential connections, but I did not speak up because I wasn't running a home server or playing games, then they came for the small business DSL customers but I did not speak up for I was not running a small business and finally they came for the corporate customers and we ended up paying thousands of dollars per server to avoid getting thrown off the 'net)...

/Mikael

Re:marketing speak = teh suck

[chrylis]

What do you mean by "non-shared"? When you get an IPv6 connection, they don't hand you a single IP address; you get a /64 or a /48, depending on the connection type.

Re:marketing speak = teh suck

[BobMcD]

I can't just decide to give my server the address 127.48.7.12 or 234.122.9.31 with IPv4, but that doesn't mean that I can't assign one within my address range.

Using NAT, you absolutely can. You're sacrificing the ability to communicate with those addresses in the wild, but that option definitely exists today.

And why exactly (1) does 'Joe Q. Sysadmin' need to select his own IP addresses and (2) can't he with IPv6?

He doesn't need to. He may want to. He has that option today.

I don't operate under the assumption that ISP's are going to hand out blocks of IPv6 addresses any more readily than they hand out IPv4's. I understand that others do. I'm not sure why they do, but since it is a futuristic sort of thing, we'll just have to wait and see. Looking at their past and present behavior, anticipating charity is dubious at best. In fact, NAT rose to popularity out of this exact same behavior. Not out of some ephemeral need to create more address space.

Re:marketing speak = teh suck

[value_added]

The point was, you don't go about memorizing them, you use DNS.

LOL. And network admins, those who are tasked with setting up and maintaining DNS, or those just doing occasional reverse lookups, do their heads just explode?

In the real world, people use IP numbers in a number of different ways, and for just as many reasons, have committed many to memory. You don't have to be a network admin, for example, to know what is behind 192.168.1.1, or that 4.2.2.1 is open for lookups.

This doesn't mean it's impossible to do the same or something similar with IPv6, of course, just that certain complaints about the complexity/awkwardness do have merit.

Re:marketing speak = teh suck

[Dog-Cow]

Where the fuck do you live where you have more than 2 viable choices for an ISP?

What universe do you live in where the "competition" would realistically compete on this feature?

Re:marketing speak = teh suck

[Anonymous Coward]

If only there were some sort of tool that could reorganize and perform computations on sets of data -- let's say, a computer -- that could store sets of data that could act as a knowledge base -- let's say, a data base -- that those network administrators could somehow use to not only store but also consult a map between the IPv6 address and any sort of info pertaining to it. One could only dream!

Re:marketing speak = teh suck

[vtcodger]

***That same competition exists under IPv4.***

Note to self. We are said have meaningful competition in digital communications that provides all sorts of benefits. Organize expedition. Try to find it/them. Surely there must be some way to profit from discovery of something that rare.

How? We can worry about how to profit once we capture one or more competes (or whatever the hell the singular of competition is) and persuade it/them to breed in captivity.

Re:thank the US government

[Chris Mattern]

US government contracts are starting to require IPv6 support.

And that's what they're getting: IPv6 support. You're getting set ups that *could* run IPv6. They don't, but they could.

Readiness test checklist

[wowbagger]

OK, here's a handy checklist to see if IPv6 is ready for prime time:

Use case: access a common web site (e.g. Slashdot) entirely by IPv6 packets:
1) Look up host's IP via IPv6 packets:
1a) Access a root DNS node via IPv6 packets (look up .org DNS server): CHECK
1b) Access .org DNS node via IPv6 packets (lookup slashdot.org address): ???
2) Access slashdot.org via IPv6 packets:
2a) Route IPv6 packets from my computer to "the Internet": FAIL
2b) Route IPv6 packets from "the Internet" to Co-Lo facility: ???
2c) Route IPv6 packets within the Co-Lo to Slashdot's servers: ???

When you (a presumably technically skilled user) can do that, then IPv6 is ready for the masses.

Re:IpV6 reality check

[r7]

The people who came up with IPv6 seemed to be too ivory tower: they forgot about
the reality on the ground. Few ISPs are even thinking about IPv6.

Amen to that. But I don't see an academic angle so much as an ILEC angle i.e., IPv6 is being handicapped by large telcos, large ISPs, legacy netblock owners and their proxies in order to drive up fees for IPv4 addresses. The threads on new fee structures, in mailing lists like arin-ppml, make this obscenely clear. IPv4 netblock owners are salivating over the potential for profit from what should be a public resource.

Only thing more disappointing than ARIN's failure to either reclaim unused IPv4 netblocks (and there are plenty of those, both large and small) or speed the adoption of IPv6 is the DOC and FCC's failure to foresee the damage, both economic and to communications, which the coming address shortage will cause.

Re:IpV6 reality check

[Daniel_Staal]

They do manifest in reality: They are why I don't have an IPv6 address: It's to much work for too little benefit. It can be worked around, but it's just more work, and wouldn't really get me anything.

Basically all he is saying is 'accept an IPv4 address as an IPv6 address'. Which would mean that 'upgrading' would be as simple as getting software that can handle being sent IPv6 addresses. (Which basically everyone's already got at this point.)

Instead at the current situation you have to figure out how and were to get an IPv6 address, and either keep an IPv4 as well (and switch between the two as the situation demands) or work out how you are going to talk to the 90+% of the world that doesn't have an IPv6 address. Either of those require extra work, for every person trying to connect to the network.

So, in the current situation, everyone who switches to IPv6 needs to be a network engineer. Because it's a complicated setup at the user's endpoint. Guess how long it'll take Grandma to switch then.

Yes, the network works, but there is no decent upgrade plan.

Re:IpV6 reality check

[Chris Mattern]

Really? Ok, then. I have a Linux box connected to a Netgear router providing NATted connections, itself connected to a cable modem that goes out to Comcast, who provides my pipe and is my ISP. Comcast ISP, by the way, does not support IPv6. If IPv6 is here and working today, I should be able to use it. How do I do that?

If you can't tell me how, than Dan's "hypothetical problems" are very real indeed.

As far as I can tell, what people have been "cheerfully ignoring" is IPv6.

Re:IpV6 reality check

[aztektum]

Grandma will upgrade to IPv6 when her ISP says your modem needs to be replaced or they have a tech swap her cable modem. The layman argument does not hold water in every situation. Most laymen will plug in their new IPv6 router and not even configure a password, let alone worry about routing tables, etc.

That's like saying grandma can't change her own brake pads, so we'll just let her grind her rotors down. Grandma will just goto a mechanic or in this case, her ISP which is staffed with NETWORK ENGINEERS. It's their fuckin' job to figure this stuff out and move to it for the benefit of their users. Your argument makes it sound like you're just a lazy network engineer who can't be bothered to work.

A pack of Luddites, honestly!

[Just Some Guy]

Every time something on IPv6 comes out, there's a thundering herd of people who've never used it but are certain that it's awful and won't ever work. What's wrong with you people? Do you feel threatened because you're used to being the networking expert among your clique and don't want to lose that reputation? If not that, then what is it that's making you sneer at a cool new technology without even trying it first?

I'm not addressing people who tried to make IPv6 work but had problems along the way, or who otherwise had bad experiences with it. That's totally understandable and I'm not going to tell such a person that they're wrong. I am talking directly to the people who've read old articles talking about why it won't work, or who are trotting out the same tired, invalid reasons to dislike it.

Here's what you need to know about IPv6:

1. It's here and working today, and a lot of people are starting to adopt it.
2. You can run IPv4 and IPv6 on the same network and machines. I don't know of any IPv6 implementation that can't run alongside IPv4.
3. DNS works perfectly fine for IPv6. I have a long address on my machines at home and work, but ever have to manually type them anywhere after adding them to DNS.
4. If you enable IPv6 alongside IPv4 and try to connect to another host, and that host has an IPv6 DNS record, then your machine will try to connect to that address and then fall back to IPv4 if that fails. If it doesn't have an IPv6 DNS record, then you'll connect via IPv4. There's no penalty for enabling it.
5. NAT sucks. It might seem like a reasonable idea until you're reminded how nice it is not to have to mess with it, then you'll come to loathe it.
6. There are plenty of good, free, reliable IPv6 tunnels available. I use Hurricane Electric, but there are lots of others to choose from.
7. All modern OSes support IPv6 out of the box.
8. Many/most consumer routers do not support IPv6 natively (although you can still tunnel through those routers from your Linux or Windows or Mac server or desktop). Some do, though, and an Airport Extreme is still a consumer product even if it's more expensive than some of the others.

I think that about covers it. There's no reason to be afraid of IPv6. If you haven't tried it, give it a shot before bragging about how smart you are for recognizing that it can't work. Again, if you've tried it and had problems, I can understand why you're leery of the idea. If you haven't at least used a free tunnel to see what IPv6 is like, though, then you don't have a lot of room to comment on the subject.

Re:ipv6experiment.com

[arkane1234]

it failed ;)