Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

IPv6 Adoption Statistics

google.com

Post reply on HN
Print view View original

71–80 of 134 posts

Re: IPv6 Adoption Statistics

#71

Post by Zancarius »

IgorPartola wrote: post #66
c0nsumer wrote: post #63

I've been trying on and off to get IPv6 working at home, but the problem I keep running into is poor performance from tunnels. I have service via Wide Open West which is great for IPv4, but they have no plans to support IPv6. So, I try using a tunnel... Both HE.net and SixXS are so incredibly slow that I get >1 second pings to something which is 30ms away via IPv4. The tunnel end point is only ~50ms away, so I can on…

Shoot an email to HE. Their support for this free service is better than most commercial support teams I've interacted with. Also, don't discount that it's possible that the other end of the equation, the server you are trying to reach, has poor IPv6 connectivity. Fire up a Digital Ocean instance for an hour (it'll cost you $0.10) and see if the site is slow from everywhere. I've been using HE.net's tunnels for a goo…

> I've been using HE.net's tunnels for a good long while now and they've been great for me.

Same here. As of about a year or two ago, my pings over the HE tunnel are only very slightly worse than those over my IPv4 interface, and I can't notice much difference in throughput. I've encountered minor issues with their tunnel servers from time to time, but usually by the time I notice the problem is resolved.

I also have to second that their support is absolutely fantastic. Plus there's always their forums, in case someone else has run into something similar. Many of tunnelbroker.net's users are equally friendly and helpful.

Re: IPv6 Adoption Statistics

#72

Post by jandrese »

c0nsumer wrote: post #63

I've been trying on and off to get IPv6 working at home, but the problem I keep running into is poor performance from tunnels. I have service via Wide Open West which is great for IPv4, but they have no plans to support IPv6. So, I try using a tunnel... Both HE.net and SixXS are so incredibly slow that I get >1 second pings to something which is 30ms away via IPv4. The tunnel end point is only ~50ms away, so I can on…

Weird. I use HE.net as well and often the latency on V6 is better than it is on V4:

My results at the moment:

ping6 google.com PING6(56=40+8+8 bytes) [Redacted] --> 2607:f8b0:4009:807::1000

16 bytes from 2607:f8b0:4009:807::1000, icmp_seq=0 hlim=58 time=22.727 ms

16 bytes from 2607:f8b0:4009:807::1000, icmp_seq=1 hlim=58 time=21.862 ms

16 bytes from 2607:f8b0:4009:807::1000, icmp_seq=2 hlim=58 time=25.147 ms

vs.

ping google.com

PING google.com (216.58.216.224): 56 data bytes

64 bytes from 216.58.216.224: icmp_seq=0 ttl=251 time=37.229 ms

64 bytes from 216.58.216.224: icmp_seq=1 ttl=251 time=36.672 ms

64 bytes from 216.58.216.224: icmp_seq=2 ttl=251 time=38.319 ms

I think this is due to my provider (Verizon) being a dick with their peering agreements and the HE.net traffic goes over a less congested pipe than my ordinary v4 traffic.

Re: IPv6 Adoption Statistics

#73

Post by mitchty »

tsmarsh wrote: post #60
wbond wrote: post #42

IPv6 became available to me on my Comcast connection in the past six months, but I ended up disabling it at my local router. Unfortunately it seems in my area (North of Boston, MA) the IPv6 routing on Comcast's network is extremely spotty. Sometimes connections would time out on all different ports (22, 80, 443). This lead to a rather poor experience for members of my household. I ran into lots of issues with SSH. My…

I found that switching to googles DNS fixed those issues for me (I'm on the south shore). There is still the occasional delay, but its similar to the ipv4 service from Comcast.

I had to switch dns for ipv4 regardless.

But ipv6 actually seems to be around 50% of my traffic now. Native dual stack is way better than the tunnel nonsense they were trialing a few years back. I've had zero issues with ipv6 and comcast now that they went to proper dual stack.

Re: IPv6 Adoption Statistics

#74

Post by IgorPartola »

Every time IPv6 comes up on HN, around 50% of the comments seem to be about how IPv6 doesn't do NAT and how now every device you have is suddenly directly exposed to the internet. Let's clarify this a bit instead of answering individual commenters:

In IPv6, just like in IPv4, you have a firewall. In Linux, you use ip6tables instead iptables, for example. This is what keeps your devices on your network safe. If you were to start from scratch to set up a router with an IPv6 firewall, you'd need just two rules: (1) allow packets in for already established connections and (2) drop every other incoming packet. If you know what you are doing, you can actually set this up yourself. I have, and while educational, it provided no real world benefit.

Most people don't want to bother with using iptables directly, so don't. Get a router that supports OpenWRT and flash it. For most of them, it's a really simple process (my TP-Link let me upload the binary to flash via the web GUI). Why OpenWRT? Well, it's secure and constantly updated, it supports IPv6 natively, and it comes with the IPv6 firewall that is configured in a fashion very similar to how you think of IPv4 (it even rate limits ping requests, etc.). As a bonus, if your ISP doesn't support IPv6, OpenWRT has an installable web GUI component for configuring an IPv6 tunnel. Lastly, even if you don't want IPv6 (yes, I see you there in the back, climbing back under your rock), still use OpenWRT. It seems to have a lot less bugs than commercial router firmware, and is a lot more stable and up to date than DD-WRT or Tomato.

Edit: One other misconception that comes up frequently is that IPv6 means that your privacy is at a more of a risk because your MAC address may be exposed. While in some configurations this can happen, IPv6 has what's called Privacy Extensions: in addition to your more permanent MAC-based IPv6 address (network prefix + munged MAC address), your OS will periodically generate a new random IPv6 address (network prefix + random number). This actually makes it marginally harder to track you since your exact IP address will change frequently, as seen by hosts you access. See http://en.wikipedia.org/wiki/IPv6#Privacy.

Re: IPv6 Adoption Statistics

#76

Post by Yeri »

JamyDev wrote: post #10
pdw wrote: post #3

Can somebody explain how Belgium achieved 28%? It's the only country that's colored bright green.

Because we pay too much for our internet :P

EDPnet, a smaller ISP has been natively providing IPv6 for ~3 years already (but 3rd biggest independent ISP?).

Telenet followed somewhere last year, and Belgacom (biggest) followed this year (only on new modems though).

Re: IPv6 Adoption Statistics

#77

Post by justincormack »

kalleboo wrote: post #54
pilif wrote: post #41

Earlier quoted context omitted.

The firewall on a local machine might suffer from exploits, thus still allowing access. Or I might want some services open to my lan and only a smaller subset opened to the public (something the personal firewalls built into many OSes can't do)

> The firewall on a local machine might suffer from exploits, thus still allowing access. Is a updated firewall from Apple, Microsoft or ipfw more or less likely to suffer from exploits than a cut-rate device from ASUS, Netgear or Linksys that hasn't been updated in years? > Or I might want some services open to my lan and only a smaller subset opened to the public (something the personal firewalls built into many OS…

The idea is that it can be turned off from the machine itself, so eg if you get hacked via a website or email, your firewall might get disabled, while another box would also need to be hacked. (Of course things like UPNP give the machines control over the router so making this moot, which is why I dodnt run them).

Re: IPv6 Adoption Statistics

#78

Post by justincormack »

IgorPartola wrote: post #66
c0nsumer wrote: post #63

I've been trying on and off to get IPv6 working at home, but the problem I keep running into is poor performance from tunnels. I have service via Wide Open West which is great for IPv4, but they have no plans to support IPv6. So, I try using a tunnel... Both HE.net and SixXS are so incredibly slow that I get >1 second pings to something which is 30ms away via IPv4. The tunnel end point is only ~50ms away, so I can on…

Shoot an email to HE. Their support for this free service is better than most commercial support teams I've interacted with. Also, don't discount that it's possible that the other end of the equation, the server you are trying to reach, has poor IPv6 connectivity. Fire up a Digital Ocean instance for an hour (it'll cost you $0.10) and see if the site is slow from everywhere. I've been using HE.net's tunnels for a goo…

They do rate limit you to around 4Mb I believe so you might want to not ship your netflix over them, which it will unless you are careful.

Re: IPv6 Adoption Statistics

#79

Post by icebraining »

iso-8859-1 wrote: post #75
cientifico wrote: post #46

In Germany, Kabel Deutschland, no longer offers ipv4. At least, My router only gets an ipv6 one. (100/6Mb + phone line = 55 eur)

How do you access IPv4 sites? Tunnel?

Probably carrier-grade NAT64.

EDIT: Apparently it's DS-Lite: https://news.ycombinator.com/item?id=6818805

Re: IPv6 Adoption Statistics

#80

Post by ay »

c0nsumer wrote: post #68
IgorPartola wrote: post #66

Earlier quoted context omitted.

Shoot an email to HE. Their support for this free service is better than most commercial support teams I've interacted with. Also, don't discount that it's possible that the other end of the equation, the server you are trying to reach, has poor IPv6 connectivity. Fire up a Digital Ocean instance for an hour (it'll cost you $0.10) and see if the site is slow from everywhere. I've been using HE.net's tunnels for a goo…

Unfortunately, it's anything that's slow... When I've got a tunnel live, Google properties and Facebook are pretty much unusable. Weirdly, sometimes it'll work fine... Other times it won't. (The server I'm testing against with is my personal site, https://nuxx.net , which has great IPv6 connectivity already. I just don't want to tunnel my home connection through it because that'll seriously push up the bandwidth use…

This description might also match a partly-working path MTU discovery (a possibly too-high rate of ICMP egress from HE end to content sites, blocked by rate-limiter on the HE device).

In IPv4 you do not notice it (it almost never triggers) because there is less tunnels and also because generally everyone does MSS clamping. In IPv6, you have the tunnel and not necessarily MSS clamping.

Two ways to tackle it:

- configure on the home router interface facing your LAN, IPv6 MTU less than you have on the tunnel (I have 1400 just because I like round numbers :-) Cleaner because works for (mostly) all protocols.

- configure the first hop router to do MSS clamping for TCP on IPv6 to 20 bytes less than what it currently does (if at all). This will work for only TCP, but that'll be the vast percentage of the traffic you are having problems with.

Post reply on HN