How important is IPv6 connectivity expected to be soon?

[Mark R]

My ISP doesn't support IPv6 at all.

Just out of interest I called a number of other ISPs in my region and they all responded the same way. They do not support IPv6 at all, and have absolutely no rollout plans for any IPv6 connectivity, including teredo within the forseable future.

Is this "normal" practice? Is it going to be of any consequence?

 

[Red Squirrel]

It is, I work for my ISP and I don't even hear anyone talk about IPv6 either. It's odd, because the last ipv4 IP range was used up a few years ago. Any new ISP would need to go IPv6 at this point or buy from another ISP.

 

[John Connor]

I'm glade IPv6 isn't that big right now. I have an old and very good WRT54GL router flashed to DD-WRT. In order for me to use IPv6 I'd have to use some broker and I don't want to mess with that. I should read more about it, but I don't see any reason to do so right now.

 

[Fardringle]

I'd say it's about as important as eating soup with a toothpick. In some places you may be allowed to try it, but it's annoying, difficult, and completely unnecessary.

 

[John Connor]

I know a guy on another forum that uses IPv6. Something to do with that IPv6 broker I was talking about. Don't know how many IPv6 addresses he resolves in a day though.

 

[apmon]

It depends a bit on the area you live in and what you do, but to some it is going to be very important.

Asia has run out of IPv4 nearly two years ago, Europe has run out nearly a year ago. In those regions it is becoming increasingly difficult to get your own IPv4 and more and more things are getting put behind carrier grade NAT. This means you can't reach these people from the outside anymore and any direct communications (not going through a central server) is quite difficult to impossible.

Therefore for many people in those parts of the world (or if you want to communicate with someone in those parts of the world) the lack of widespread support of IPv6 is a real problem and results in a real loss of functionality already today. E.g. you can't access your local computer anymore for file transfer or remote administration or peer to peer communication.

Many people now a days simply put everything "into the cloud" and that usually works through NAT. But there are also many people who care about their privacy and controlability and like to keep their data local. For them the lack of IPv6 support is a big issue!

In order to preserve the nature of the Internet as a network where anyone can communicate with anyone, moving to IPv6 is therefore extremely important. IPv6 won't suddenly allow you to do new things, but it will prevent the loss of old functionality. Unfortunately that is not very "sexy" and so the transition has been extremely slow :-(

 

[Mixolydian]

I for one am very happy with IPv6. Finally puts an end to the NAT bullshit. Since each device on the network has a globally-unique IPv6 address, I can access any device outside my network without any configuration changes.

 

[Red Squirrel]

I hate how with ipv6 they changed everything, and that it includes the mac address, why include such redundant information and waste space? They should have made it work the exact same way as ipv4 but just add more octets. I'm also not a big fan of the fact that they got rid of NAT. NAT is important in drawing a line between the public internet and your own private network IP space. If every device on your network has to have a public IP address then you have to bank on the fact that your range wont change, which it probably will if it's a residential service or if you change ISPs or have two ISPs with one as backup. The other thing that gets brought up is security, but that is not too much an issue if you manage your firewall properly, though the average user wont even have a firewall.


I've been wanting to setup a lab to play with it though. I believe Hurricane Electric will actually give you a routable range for free to play with.

 

[Mixolydian]

I hate how with ipv6 they changed everything, and that it includes the mac address, why include such redundant information and waste space? They should have made it work the exact same way as ipv4 but just add more octets. I'm also not a big fan of the fact that they got rid of NAT. NAT is important in drawing a line between the public internet and your own private network IP space. If every device on your network has to have a public IP address then you have to bank on the fact that your range wont change, which it probably will if it's a residential service or if you change ISPs or have two ISPs with one as backup. The other thing that gets brought up is security, but that is not too much an issue if you manage your firewall properly, though the average user wont even have a firewall.


I've been wanting to setup a lab to play with it though. I believe Hurricane Electric will actually give you a routable range for free to play with.

Yes the address format is far more complicated, which is why I use DNS to manage IPv6 addresses.

 

[Cooky]

If you're in N. America, don't expect the residential ISP's to provide IPv6 anytime soon.
To support IPv6, carriers need to spend more money on equipment & manpower.
I don't see any benefits for them to do that.
Comcast was an exception - I think that was more a marketing campaign than anything else.

 

[cmetz]

My ISP doesn't support IPv6 at all.

Just out of interest I called a number of other ISPs in my region and they all responded the same way. They do not support IPv6 at all, and have absolutely no rollout plans for any IPv6 connectivity, including teredo within the forseable future.

Is this "normal" practice? Is it going to be of any consequence?

This is normal.

If there was a real market demand, ISPs would eventually offer it. The only demand is from people who have convinced themselves that IPv6 is the next big thing and they want to have the new thing so they can be all ahead of the curve.

IPv6 is now over 20 years in the making. There comes a time when you need to say okay, had its chance, FAILED, now what are we really gonna do?

 

[apmon]

If you're in N. America, don't expect the residential ISP's to provide IPv6 anytime soon.

Perhaps not a residential ISP, but if I am not mistaken, Verizon uses IPv6 for its LTE network. T-Mobile is also working on using IPv6 in its entire network and you can already test it if you want.

About 3% of traffic from the US to Google now uses IPv6. In Switzerland it is now at 10% of all internet users using IPv6. Not a huge amount yet, but definitely progressing in the right direction.

 

[apmon]

If there was a real market demand, ISPs would eventually offer it.

It is a typical failure of markets or what is called the "tragedy of the commons".
Because of that, things will continue to increasingly break and get worse. But because no one benefits from taking the first step, nothing happens even though everyone agrees that it is the correct thing to do.

 

[Red Squirrel]

Yes the address format is far more complicated, which is why I use DNS to manage IPv6 addresses.

Yeah guess it wont be so bad given you can just use DNS for everything. Heck even with ipv4 I use DNS at home. I only know a few of my servers' IPs by heart. Though the lack of NAT would still make it so every time your range changes you have to go edit your DNS records.

I'm guessing at some point NAT will make a come back for ipv6 as it will be required for businesses that have 1000's of computers and can't afford to not manage the IP addressing. Perhaps it would be some kind of 1:1 NAT.

 

[apmon]

Though the lack of NAT would still make it so every time your range changes you have to go edit your DNS records.

Internally you can use the link local address. That way you can communicate with all of your devices behind your router without having to worry if the external global prefix changes.

If you have multiple routers in your network topology, e.g. because you are a company with 1000s of PCs, you can use site local addressing. Again, in that case you don't have to worry about renumbering internal numbers.

But probably the best way is to automate the update of the dns record when the IP changes.

 

[drebo]

Using DHCP (or stateless autoconfigure or EUI-64) and DNS, you can certainly get around the issue of "renumbering" your network should your private space need to change for some reason. Indeed, the network discovery protocol was more or less designed with that in mind and makes it quite simple to change scopes.

If your facility is large enough, you would simply buy an allocation and then you don't need to worry about it. For a small business or residence, if the network was properly put together, it wouldn't matter.

 

[Red Squirrel]

Internally you can use the link local address. That way you can communicate with all of your devices behind your router without having to worry if the external global prefix changes.

If you have multiple routers in your network topology, e.g. because you are a company with 1000s of PCs, you can use site local addressing. Again, in that case you don't have to worry about renumbering internal numbers.

Hmm how would that work? Like say you do want one of those link local to listen on a port. Sounds like that could be the solution to keeping internal IPs the same, and under some form of control.

Also raises another point if you were to depend on the public IP range and your internet goes down would you still be able to talk to local devices?

 

[VirtualLarry]

Comcast was an exception - I think that was more a marketing campaign than anything else.

Marketing or not, Comcast is actually doing an IPv6 rollout for most of their broadband customers. Proving that they, not Verizon, have the "most advanced network".

I'm still stuck on IPv6-less FIOS. 🙁

 

[VirtualLarry]

IPv6 is now over 20 years in the making. There comes a time when you need to say okay, had its chance, FAILED, now what are we really gonna do?

IPv6 failed, you say? Ok then, what about the fact that there are twice as many IPv6-connected nodes as last year.

And if IPv6 is a failed experiment, I shudder to think of the alternatives... carrier-grade NAT for everybody? No true peer-to-peer internet anymore?

To me, the failure of IPv6 would be the death of the internet as we know it. I really don't want that to happen.

 

[drebo]

Hmm how would that work? Like say you do want one of those link local to listen on a port. Sounds like that could be the solution to keeping internal IPs the same, and under some form of control.

Also raises another point if you were to depend on the public IP range and your internet goes down would you still be able to talk to local devices?

The idea is that you use the link-local address for internal communications. That way, it doesn't matter what the IP address of the device is in terms of external communications, as your internal address is always the same. The drawback, of course, is potentially massive broadcast domains.

In regards to public IPs when the Internet goes down, that doesn't matter. It'd react the same way as it currently does. Just because the rest of the world doesn't know about your subnet (for whatever reason) does not mean that subnet does not, or cannot, exist. Locally, it still exists.

 

[drebo]

IPv6 failed, you say? Ok then, what about the fact that there are twice as many IPv6-connected nodes as last year.

And if IPv6 is a failed experiment, I shudder to think of the alternatives... carrier-grade NAT for everybody? No true peer-to-peer internet anymore?

To me, the failure of IPv6 would be the death of the internet as we know it. I really don't want that to happen.

Carrier NAT doesn't happen in the US outside of shitty WISPs, and it won't happen on a large scale.

There are still unassigned /24s available from ARIN. There are also unused blocks that can be reclaimed. I worked for a university for a while that had a /16 and assigned a static public IP to every single node within the network. Forcing people who were over-allocated to clean up their networks and be responsible will add years to IPv4.

Also, when people talk about IPv4 exhaustion and that all IP blocks have been used, they're being intentionally misleading. IANA has allocated all blocks to RIRs, yes. However, the RIRs still mostly have available blocks to assign to end users who need them...yes, even RIPE and AFRINIC. If you can justify it, you can still get a /24 or potentially larger block in the US. ARIN has made justification more rigorous, but anyone can still do it.

 

[apmon]

Carrier NAT doesn't happen in the US outside of shitty WISPs, and it won't happen on a large scale.

Yes, the US is still in the fortunate situation that it has some addresses left. But the US is not the only country on this planet. In fact it is only about 5% of this planet (by population). And even the US is projected to run out of IPv4 addresses in less than a year (probably sometime in the first half of 2014). And that is despite most mobile devices already being behind CGN.

There are still unassigned /24s available from ARIN. There are also unused blocks that can be reclaimed. I worked for a university for a while that had a /16 and assigned a static public IP to every single node within the network. Forcing people who were over-allocated to clean up their networks and be responsible will add years to IPv4.

Asia burnt through 25 /8 in 3 years. Those couple of underused /16 aren't going to get you very far. Even reclaiming some of those under used /8s is only going to gain you a couple of more months (and we have had about 20 years to address the issue and failed misserably) so that isn't going to change anything. Furthermore, reclaiming (and renumbering) those underused space is possibly going to cost as much as moving over to IPv6 for little to no gain.

Also, when people talk about IPv4 exhaustion and that all IP blocks have been used, they're being intentionally misleading. IANA has allocated all blocks to RIRs, yes. However, the RIRs still mostly have available blocks to assign to end users who need them...yes, even RIPE and AFRINIC.

Wrong! IANA ran out of addresses in Feb 2011. APNIC (Asia) then depleted in April 2011 only 2 months later. RIPE (Europe and the Middle East) is also depleted since Sep 2012. As a new customer (ISP) you can still get a /22. But good luck trying to run an ISP with 1000s of customers if you only have 1024 addresses available.

I don't know what the situation is like in Asia, but in Europe you are starting to see residential ISPs having to put customers behind CGN. And it is costing them dearly in customer dissatisfaction because various peer-to-peer services like gameing no longer work as they used to. Unfortunately without everyone else (who isn't experiencing the issues yet) supporting IPv6, there is little they can do about that customer dissatisfaction.

But eventually everyone will have a worse off experience, because no one wanted to move.

If you can justify it, you can still get a /24 or potentially larger block in the US. ARIN has made justification more rigorous, but anyone can still do it.

That will likely change in mid 2014 once ARIN has also run out of addresses. But who could possibly be expected to plan a whole 12 month in advance... ;-)

 

[theevilsharpie]

I hate how with ipv6 they changed everything, and that it includes the mac address, why include such redundant information and waste space?

Because it allows stateless autoconfiguration with a permanent address, which is something that IPv4 couldn't do.

I'm also not a big fan of the fact that they got rid of NAT. NAT is important in drawing a line between the public internet and your own private network IP space. If every device on your network has to have a public IP address then you have to bank on the fact that your range wont change, which it probably will if it's a residential service or if you change ISPs or have two ISPs with one as backup.

IPv6 has Unique Local Addresses exactly for this use case.

IPv6 is now over 20 years in the making. There comes a time when you need to say okay, had its chance, FAILED, now what are we really gonna do?

There has been no market demand for IPv6 among end-users because IPv4 have traditionally been so cheap. Also, support for IPv6 has ranged from "mostly works" to "what's an IPv6?," so many companies still consider it experimental and are leaving it to others to trial before jumping on board.

Once the price of IPv4 addresses begins increasing, expect to see more organizations implement dual-stack environments.

 

[Mushkins]

Once the price of IPv4 addresses begins increasing, expect to see more organizations implement dual-stack environments.

That was kinda my take on it. All the people who insist they need NAT for whatever reason, just... run IPv4 internally but IPv6 on the WAN connections and let the router handle translations. Hell, most consumer grade $50 routers support this sort of configuration these days, even though nobody is really using it. Until the network hardware stops supporting IPv4 entirely (not likely considering how many ancient protocols are still supported by every router ever) there is really no rush to switch your internal network to v6 exclusively just because the ISPs are pushing it.

Yeah, there's a little more routing overhead, but for 99% of use cases it's a total non-issue. For that other 1% of companies, odds are "buy some more routers and do better load balancing/traffic management" is the easier solution than "redesign and readdress the whole network for IPv6" anyway.

 

[ScottMac]

I believe one of the reasons for the slow roll-out is related to the equipment. Because the addresses are so much larger it takes more ram, more ROM/EPROM/EEPROM, and more execution space. Devices that are slow now would be brought to it's knees but using IPv6. Firewalls / firewall processes are much more difficult, and would likely take more time to accomplish.

Most ISPs (if not all, by now) are IPv6-capable in the core, and many/most/all offer IPv6 to their commercial customers. Updating / converting their access devices will likely be by natural evolution, replacing the old and obsolete with the newer IPv6-capable equipment.