The cable modem change I did yesterday appears to be a succes. I now
have stable native IPv6. Dual stack.
So far so good...
The cable modem change I did yesterday appears to be a succes. I now
have stable native IPv6. Dual stack.
The cable modem change I did yesterday appears to be a succes. I
now have stable native IPv6. Dual stack.
Great!
Ziggo??
The cable modem change I did yesterday appears to be a succes. I
now have stable native IPv6. Dual stack.
So far so good...
Did you switch providers also then?
... "Не знаю. Я здесь только работаю."
So far so good...
Did you switch providers also then?
So I just changed the cable modem for a model that supports IPv6.
As I said, they give you the CPE on loan. The contract states that it
has to be returned eventually. With a penalty of EUR 65 for failure.
So these modems are not supposed to apeear on the market. Fact is that they do. Aldo they have the legal right to demand return of the ol
modes, they often do not insist and customers are requested to dispose
of old modems themselves. Also modems get lost because of
adminstrative errors.
The second gamble was if I could get it registered with- and accepted
by the cable company. I didn't get it through their official channels. From their POV it is a renegade. Much to my surprise however I managed
to register it with the activation codes supplied to me for my
previous modem.
So now I am a happy camper...
So far so good. It does what I want it to do. I have stable dual
stack. But I have already found some quirks.
To be continued...
... "Не знаю. Я здесь только работаю."
I left that behind... I am pensioado. ;-)
I now have stable native IPv6. Dual stack.
The cable modem change I did yesterday appears to be a succes. I
now have stable native IPv6. Dual stack.
Great!So it seems. But I am not cancelling my tunnel accounts yet...
Ziggo??Yes. They are rolling out IPv6, but it is tantalizingly slow.
I have been nagging them for over five years. After some failed
attempts they finally seemed to get going earlier this year, but after
a promising quick rise, it flatttened out again. :( http://www.worldipv6launch.org/apps/ipv6week/measurement/images/graphs /Ziggo.png
Presumable all the CMTSs are ready, the bottle neck is in the CPE.
So I just changed the cable modem for a model that supports IPv6.
Did you know this needed to be done previously? Or did something tip
you off that Ziggo was indeed rolling out IPv6 but you were unable to utilize it?
As I said, they give you the CPE on loan. The contract states
that it has to be returned eventually. With a penalty of EUR 65
for failure. So these modems are not supposed to apeear on the
market. Fact is that they do. Aldo they have the legal right to
demand return of the ol modes, they often do not insist and
customers are requested to dispose of old modems themselves. Also
modems get lost because of adminstrative errors.
Our cable company also goes by this method. Although I do think we're allowed to buy our own modems. With the ease of them providing one for you, you don't ever feel the need to do so until you actually analize
your bill and see how much you've paid for the same modem in rent over
the last 2-3 years. I could have easily bought and paid for a couple already.
The second gamble was if I could get it registered with- and
accepted by the cable company. I didn't get it through their
official channels. From their POV it is a renegade. Much to my
surprise however I managed to register it with the activation
codes supplied to me for my previous modem.
Understood. It's pretty surprising that they didn't notice that modem
had already been registered by someone else.
So now I am a happy camper...
That's what matters in the end.
So far so good. It does what I want it to do. I have stable dual
stack. But I have already found some quirks.
As you will, I'm sure.
Sometimes it seems like the people working on introducing IPv6 at the providers know about as much if not less about it than we do. I was
told I'd get an email (along with everyone else) letting me know when
IPv6 was functional. Well, still no email.. but I've been using native IPv6 for some time already.
So it seems. But I am not cancelling my tunnel accounts yet...
Sixxs has done it for me ;)))
Congrats!
Sixxs has done it for me ;)))
Sorry to say this, but Sixxs are a bunch of morons anyway.
When I tried to register for a tunnel, they've rejected my request because in their opinion I got no reliable email address running my
own MTAs for longer than Sixxs exists. <facepalm>
Basically the cable modem just needs to support the right DOCSIS
version and its MAC has to be activated, i.e. the cable provider puts
the MAC into the database and links it to your customer ID. There's no authentication like it's done for DSL (PPP).
So it seems. But I am not cancelling my tunnel accounts yet...
Sixxs has done it for me ;)))
But HE still works great.
http://www.worldipv6launch.org/apps/ipv6week/measurement/images/g
raphs /Ziggo.png Presumable all the CMTSs are ready, the bottle
neck is in the CPE.
Hmmm .. actually slightly going downwards again ... strange.
Not allways the CPEs (I know of CPEs that have been tested 4 years
ago, and that would fully comply, but they are not reconfigured to use IPv6 :( )
Coming to think of it, they never asked me to return my 10+ year old Motorola modem either...
Here we do not have that choice. There is no rent for the modem
specified on the bill. Officially you do not rent it, it is given on
loan "for free". I presume this is a legal construction that allows
them to force you to use their modem. If they would rent it out for a
fee, the customer could refuse.
Like you, I would very much prefer the German model. Not just because owning your own stuff usually is cheaper in the long run but also
because the stuff from the providers is not all that good. There is
much better equipment to be had on the open market. I'd go for a
Fritzbox cable modem if I has the choice.
They can see the MAC, so they should be able to see that it jas been registered before. But they do re-issue refurbished modems, so seeing
a previously issued modem turn up at another customer is not an error condition per sé.
Like that the web interface of the modem is not reaceable via IPv6. So this box is not usable in an IPv6 only environment...
So.. in your case it is not "official" either. It can go away
tomorrow...
Coming to think of it, they never asked me to return my 10+ year
old Motorola modem either...
As long as they're not charging you for it. Enjoy your new 10+ year
old modem! :)
Like you, I would very much prefer the German model. Not just
because owning your own stuff usually is cheaper in the long run
but also because the stuff from the providers is not all that
good. There is much better equipment to be had on the open
market. I'd go for a Fritzbox cable modem if I has the choice.
I haven't done enough studying on modems to have an opinion on what's
good or not. I currently have a Motorola Surfboard SB6141 which seems
to have all the bells and whistles I currently need and/or want.
I would probably have bought a new modem on my own if it were to give considerably better speeds, but I'm currently paying for 50mb/5mb, and speedtest.net reports 60/6.. so I can't really complain. :)
Like that the web interface of the modem is not reaceable via
IPv6. So this box is not usable in an IPv6 only environment...
Are you trying to connect to your fe80:* address? Or the actual
address given by your provider?
So.. in your case it is not "official" either. It can go away
tomorrow...
Exactly. And like you, I have just disabled my he.net tunnel in case anything were to go wrong I won't be without IPv6 for long.
Sixxs has done it for me ;)))
Sorry to say this, but Sixxs are a bunch of morons anyway.
When I tried to register for a tunnel, they've rejected my request
because in their opinion I got no reliable email address running my
own MTAs for longer than Sixxs exists. <facepalm>
When I tried to register for a tunnel, they rejected my request
because they could not locate my street address.
I replied with google maps link.
They did not answer.
The following comes to mind. I have been assigned a /56. So I can have
256 Iv6 subnets. But... I have not as yet discovered a way to use
anything but the first /64 of that assigment with my modem/router. How about you?
You get that speed on both IPv4 and IPv6?
Like that the web interface of the modem is not reaceable via
IPv6. So this box is not usable in an IPv6 only environment...
I have tried them all. The web interface only responds at IPv4. No big deal. Yet... But of course in future equipment should be manageable
via IPv6.
When I tried to register for a tunnel, they rejected my request
because they could not locate my street address.
I replied with google maps link.
They did not answer.
I have no idea what my ISP assigned me,
though quick searches of the TWC forums seem to mention a /56. But it seems my router is only giving me a /64 as well. Granted, I'm using
the "Native" connection type, and also using DHCP-PD. If I disable
either one of the two, I am able to input my own data and request different sized prefix lengths.
Since I don't really see a reason at the moment to need 256 subnets
for less than 20 devices.. I probably won't dive into it much further
at this time.
You get that speed on both IPv4 and IPv6?
Looks like my IPv6 speed test is a little bit faster. 61.2mb/s (IPv4)
vs. 62.5mb/s (IPv6). Not really much of a difference, but looks like I
do indeed get that speed on both. This was also tested on ipv6-test.com/speedtest and the nearest testing location for me was in
the UK.
Like that the web interface of the modem is not reaceable via
IPv6. So this box is not usable in an IPv6 only environment...
I don't think I've ever even tried this. While my router has it's own
IPv6 address, I can't seem to connect to the router's web config via
IPv6 either.
I have tried them all. The web interface only responds at IPv4.
No big deal. Yet... But of course in future equipment should be
manageable via IPv6.
I'm sure it will. But as long as the router handles dual stack, it's
most likely not an issue.
If "please.send.me.somewhere.cool" was a viable address, even that
could be used to access your router inside your LAN.
Free service or not, he was acting like a moron.
Does your cable modem/router got several Ethernet ports or just one?
Does it support VLANs? If the answer is "one" and "no", you'll need another router for that.
BTW, we had those "Zwangsrouter" too for some telcos, but it became history a few months ago. Luckily!
PD ;) Try a DHCP client supporting prefix delegation and ask the cable modem/router for a prefix.
I have no idea what my ISP assigned me,
I can see it on the setup page of the modem/router:
Network Configuration
IPv6 Address: 2001:1c02:1100:fd00:7277:81ff:xxxx:xxxx/64
IPv6 Prefix: 2001:1c02:1100:fd00::/56
IPv4 Address: 192.168.178.1
MAC Address 70:77:xx:xx:xx:xx
Interface/Prefix
LAN Delegated Prefix 2001:1c02:1100:fd00::/64
WAN
IPv6 Address: fe80::7277:81ff:xxxx:xxxx/64
2001:1c02:1100:0:815e:4c13:xxxx:xxxx/128
IPv6 MTU Size: 1500
IPv4 Address: 83.85.196.65
MAC Address: 70:77:xx:xx:xx:xx
Duration D: 07 H: 00 M: 00 S: 00
Expires: Mon Dec 05 14:23:29 2016
IPv4 DNS Servers: 89.101.251.228
89.101.251.229
IPv6 DNS Servers: 2001:b88:1002::10
2001:b88:1202::10
2001:730:3e42:1000::53
WAN Connection Type: DHCP
I x't out part of some of the addresses and MACs. The only thing I can change here is the IPv4 local address (192.198.178.1).
As you can see, the prefix assigned to me is a /56. From the /56, the first /64 is assigned to the LAN by prefix delegation.
The HTML interface BTW is the only way to access thr outer. No SSH interface...
though quick searches of the TWC forums seem to mention a /56.
But it seems my router is only giving me a /64 as well. Granted,
I'm using the "Native" connection type, and also using DHCP-PD.
If I disable either one of the two, I am able to input my own
data and request different sized prefix lengths.
Then you can do more than I can.
Since I don't really see a reason at the moment to need 256
subnets for less than 20 devices.. I probably won't dive into it
much further at this time.
Same here and I have less than 10 devices connected. I do not need
more han a /64. Presently. I say "presently" because it it IPv4 think. IPv6 was designed to give "end connections" the ability to devide
their network into subnets. Subnets that can be isolated and
firewalled off from each other.
What about a guest subnet? You want your guests to have full internet access, but you do not want them in the same subnet as the computer
you use for banking do you? And your internet of things.. You want
those smart devices in a seperate subnet. Same for your car when it is
in dock....
Your provider does not have a speedtest of its own?
I don't think I've ever even tried this. While my router has it's
own IPv6 address, I can't seem to connect to the router's web
config via IPv6 either.
How about SSH?
I'm sure it will. But as long as the router handles dual stack,
it's most likely not an issue.
Not yet. But it will be an issue in the future. And I don't understand
why they didn't make the web interface dual stack right away. HTML
over IPv6 isn't rocket science. It is usually easier to do such things right away from scratch than to have to add it later.
That probably illustrates one of the propblems with IPv6. Lots of developers are still stuck in IPv4 think. IPv6 is added as an afterthought...
If "please.send.me.somewhere.cool" was a viable address, even
that could be used to access your router inside your LAN.
Some might want to access it from outside. (Not me..)
PD ;) Try a DHCP client supporting prefix delegation and ask the cable modem/router for a prefix.
PD ;) Try a DHCP client supporting prefix delegation and ask the
cable modem/router for a prefix.
I think that's exactly what I was referring to. If I disable DHCP-PD
on my router, I can indeed ask for a prefix and/or prefix length. I
just haven't tried it yet, and don't feel like breaking anything at
the moment since everything is working so nicely. :)
I think that's exactly what I was referring to. If I disable
DHCP-PD on my router, I can indeed ask for a prefix and/or prefix
length. I just haven't tried it yet, and don't feel like breaking
anything at the moment since everything is working so nicely. :)
It's a prefix tree. The main router connected to the WAN link uses
DHCP PD to request a prefix from the provider. Some routers allow to specify the prefix length. Usually the provider will assign a /56 to
your router. After that your router takes the first /64 from that /56
and uses it for its LAN interface. Again, some routers allow to modify
the prefix length for the LAN. When you connect another router to your main router, the new router can ask for a prefix via DHCP PD. The main router consults its table and assigns the second /64 out of the
provider's /56. This also implies that each router assigning a prefix
to another one also takes care about routing the prefix to the correct next hop i.e. the new router.
Thanks for the detailed explanation. That helps me understand the workings quite a bit better. I can only seem to modify the prefix
Also, I noticed my WAN IPv6 address has a different prefix than my
LAN prefix. Which seems like it would fit into the category you're describing, right?
All of this could probably be done via one router, but it would
probably need something more advanced than what they're providing in stock firmware (ie: OpenWRT). With stock firmware, it seems it would
be easier to setup different subnets with separate access points or routers added to the mix, which would make for a pretty cool project
with a couple extra Raspberry Pi's.
The RasPi is a bad choice for a router, since the LAN is actually USB based.
he.net is fair and reasonable, SixXs is wayward. But it doesn't matter anymore, since the latter doesn't accept any new users.
And hopefully everyone will get native IPv6 in the forseeable future.
Does the cable modem/router allow to separate a single Ethernet port,
i.e. creating an additional LAN?
So we won't open that can of worms :)
There was some support by a few manufacturers, but most pressure came
from the public, and there's also EU directive 2008/63/EC.
Yes.
Not yet. But it will be an issue in the future. And I don't
understand why they didn't make the web interface dual stack
right away. HTML over IPv6 isn't rocket science. It is usually
easier to do such things right away from scratch than to have to
add it later.
I'm not sure what the issue(s) would be in the future. As long as the router itself supports dual stack, it's only a LAN IPv4 address, which wouldn't matter if you actually had functional IPv4 outside your LAN
or not. Obviously if you wanted to access your router from outside the
LAN (not me), it would present a problem.
That probably illustrates one of the propblems with IPv6. Lots of
developers are still stuck in IPv4 think. IPv6 is added as an
afterthought...
Hasn't that always been the case (IPv6 being an afterthought)? At the
time IPv4 was created, they had no idea they would ever need something more.
If "please.send.me.somewhere.cool" was a viable address, even
that could be used to access your router inside your LAN.
Some might want to access it from outside. (Not me..)
Agreed. My router would never be open to the public.
The RasPi is a bad choice for a router, since the LAN is actually USB based.
Very good point. The RasPi only has an incoming ethernet port, no outgoing.
There is that, but I was thinking a bit further ahead. Dual stack is
not the end goal, it is also a transition mechanism. A sensible
engineer does not want to maintain duality forever. On the contrary,
once you got the new syetem in place, you want to get rid of the old system ASAP. Keeping two systems in the air side by side increases the effort of maintenace and increases the chanco oe errors. Mars landers
and planes have crashed because of using metric and the imperial
measures in the same project.
IPv4 only stuff is an obstacle on the road to IPv6 only.
That was than, but just like the first cars were designed to function
is a horseless society, IPv6 was designed to operate in an environment without IPv4.
Agreed. My router would never be open to the public.
"Open to the public" and accesible from outside is not the same thing.
YOU may want to access it from outside, but not giving that access to anyone else.
Very good point. The RasPi only has an incoming ethernet port, no
outgoing.
That sentence doesn't really make sense.
It's no problem to run more than one network on top of a single
ethernet port, either in nasty ways by just overlaying them or with
VLANs or tunnels. But the issue mentioned first, it being on the far
end of a USB bus means it will have issues with high traffic rates,
which makes it somewhat unsuitable for being a router on modern
networks.
Maybe not, but I know what I was saying. My current router has an incoming ethernet port and 4 outgoing ethernet ports. The RasPi only has one ethernet port.
[..]There is that, but I was thinking a bit further ahead. Dual stack
is not the end goal, it is also a transition mechanism. A
sensible engineer does not want to maintain duality forever. On
the contrary, once you got the new syetem in place, you want to
get rid of the old system ASAP. Keeping two systems in the air
You may be thinking much to far ahead (which is not a bad thing). It
has taken this long to get IPv6 deployed. I can only imagine how long
it's going to take to get rid of IPv4 completely.
IPv4 only stuff is an obstacle on the road to IPv6 only.
True. But their thought process probably had something to do with the
life expectancy of said router(s) would probably not outlast the transition to full IPv6, and while it's in transition and even
completes, the router would still work fine, without extra work put
into it.
You know the routine.. "sell for the highest price possible, but put
the least amount of effort into the product" kind of deal.
That was than, but just like the first cars were designed to
function is a horseless society, IPv6 was designed to operate in
an environment without IPv4.
True. Albeit I'm sure they've all been brainwashed into the "we've got plenty of time" mindset already.
Agreed. My router would never be open to the public.
"Open to the public" and accesible from outside is not the same
thing. YOU may want to access it from outside, but not giving
that access to anyone else.
I wouldn't want my router accessible from outside whatsoever. Anything
I need to do I can do from within my LAN.
If my router were to go down while I wasn't home, there's most
likely no way of fixing it from there anyways.
My main firewall/router/server has one Gigabit Ethernet port. That
port is connected to a switch that supports VPN's and VPN trunking.
Als Linux supports VPN and trunking, I use a trunk between the Server
and teh Switch, the switch is trunked to a second smart switch.
I have used a Pi as a router in my motorhome. A power-Wifi or a
3g-dongle provide a Wan connection. A small usb Wifi then provides the in-vehicle Wifi. Of course the Lan port is also available. This setup
can cope with the slow speeds of the wan connections. I intend to use
a Pi3 next year, then the internal Wifi can give the in-vehicle lan.
You may be thinking much to far ahead (which is not a bad thing).
It has taken this long to get IPv6 deployed. I can only imagine
how long it's going to take to get rid of IPv4 completely.
Maybe not as long as you think. Some apparently are already seeing the light when it comes to the advantages of IPv6 only internal networks:
http://preview.tinyurl.com/hrez4e9
The IPv6 train is slow in coming on steam. But when it starts gaining momentum, and when we see IPv6 only internal networks, it may
accelerate expom\nentially.
True. But their thought process probably had something to do with
the life expectancy of said router(s) would probably not outlast
the transition to full IPv6, and while it's in transition and
even completes, the router would still work fine, without extra
work put into it.
And missing the opportunity to gain experience when it is not critical yet. Fixing the roof when the sun shines...
You know the routine.. "sell for the highest price possible, but
put the least amount of effort into the product" kind of deal.
Good short term strategy, but a company that does not invest in the
future is heading for a dead end.
True. Albeit I'm sure they've all been brainwashed into the
"we've got plenty of time" mindset already.
I think at least some are coming to see the light. It may be different
in your part of the world, but here it is very difficult for small companies to get IPv4 address space. The industry is reaaly starting
to feek the pain of IPv4 exhaustion.
Same for me, but not everyone is in that position. We have sysops here
who only have physical access to their systems in the weekend. OTOH,
there are pensionados who are gone for a couple of month in the
summer. Roaming the world in a camper, or playing with boats..
You do not have a neighbour or friend who can push the reset button?
So it seems. But I am not cancelling my tunnel accounts yet...
Sixxs has done it for me ;)))Sorry to say this, but Sixxs are a bunch of morons anyway.
So it seems. But I am not cancelling my tunnel accounts yet...
Sixxs has done it for me ;)))Mine is still working but SixXs has stopped accepting requests for new tunnels
and Jeroen Massar shows signs of becoming tired of the
project, so I think we should not expect much from SixXs any more.
But HE still works great.It may not last forever either...
Hmmm .. actually slightly going downwards again ... strange.Not really considering that there is only one model of modem that
supports their IPv6 and that it only provides dfor 8 bonded channel,
which is not enough for the highest speed Ziggo offers. (300/30) So if someone upgrades to the highest speed, he/she gets a modem than does
16 bonded channels. But without IPv6. Same when a modems breaks.
Michiel van der Vlist wrote to All <=-
The cable modem change I did yesterday appears to be a succes. I now
have stable native IPv6. Dual stack.
Markus Reschke wrote to Nicholas Boel <=-
With OpenWrt you can separate Ethernet ports (if supported by hardware) and WLANs, while giving each one its own prefix. This is very useful
for things like a guest WLAN.
be easier to setup different subnets with separate access points or routers added to the mix, which would make for a pretty cool project
with a couple extra Raspberry Pi's.
The RasPi is a bad choice for a router, since the LAN is actually USB based.
Have you called your ISP? If so what was their response?
:f1d0:2:221:6 (2:221/6)
Have you called your ISP? If so what was their response?
I asked once. The answer was something like "and how would it make
your life better with the ipv6 support in your adsl line?"
Unfuckingbelievable.
The IPv6 train is slow in coming on steam. But when it starts
gaining momentum, and when we see IPv6 only internal networks, it
may accelerate expom\nentially.
These exact same things were said when IPv6 was first introduced.
Even "World IPv6 Day" was what, 4 years ago?
It's still not fully implemented yet.
So while I am indeed excited about the movement, I'm not going to hold
my breath for exponential accelleration. :)
True. But their thought process probably had something to do
with the life expectancy of said router(s) would probably not
outlast the transition to full IPv6, and while it's in
transition and even completes, the router would still work fine,
without extra work put into it.
And missing the opportunity to gain experience when it is not
critical yet. Fixing the roof when the sun shines...
Yep. Unfortunately that's how the world works.
I think at least some are coming to see the light. It may be
different in your part of the world, but here it is very
difficult for small companies to get IPv4 address space. The
industry is reaaly starting to feel the pain of IPv4 exhaustion.
Sure, those smaller companies are probably the first to deploy newer technologies, while the bigger companies with monopolies hold off
until they can not hold off any longer.
Mine is still working but SixXs has stopped accepting requests
for new tunnels
Indeed ... and they auto disabled my tunnel after a PoP went offline,
and refused to reactivate it ...
and Jeroen Massar shows signs of becoming tired of the
project, so I think we should not expect much from SixXs any
more.
Indeed
But HE still works great.
It may not last forever either...
I cannot influence it. Currently it works, and if it stops, I need to reevaluate.
Strange ... in Austria they only offer IPv6 capable CPEs since some
time now. The only way to avoid DS-Lite is to use an older CPE, which
are not offered any more.
I have a Ubee Bridge with DOCSIS 3.0 and no WiFi. If it breaks down I
get a new DOCSIS 3.1 CPE with WiFi and no public IPv4 any more ...
:f1d0:2:221:6 (2:221/6)
I must say that I just love the above. Beautiful!
Even "World IPv6 Day" was what, 4 years ago?
First there was "World IPv6 day". That was in june 2011. Followed a yearlater by "IPv6 launch day" on 6 June 2012. Many content providers switched on IPv6 then.
It's still not fully implemented yet.
There is progress. I have an add-on intstalled for Firefox that show
me when a website is connected over IPv6.
https://addons.mozilla.org/nl/firefox/addon/ipvfox/
I find it encouraging how many content providers already support IPv6.
I am not going to hold my breath either, but I think we will
eventually see the train coming on steam. It is unavoidable.
Yep. Unfortunately that's how the world works.
It appears I was wrong. The Ubee modem/router DOES have IPv6 access to
its web interface. But... not from a Win Xp system. On my laptop
running Win 7, I can access the web interface via the link local
address,
Not all of the bigger ones and not in all countries. But some of the
big ones ARE moving. In the US, Comcast, Verizon and T-Mobile ARE
moving:
Seems to be part of their exit strategy. No new users, no new tunnels
and when a POP provider stops, the tunnels running through that POP
are terminated.
There are 15 nodes in IPV6.IP6 running an he.net tunnel. If he.net
were to go as well, that would heva a serious impact on the Fidonet
IPv6 vommunity. Fortunately I see no signs of he.net handing in their dinner pail yet...
I'm keeping my eye out for an IPv6 DNS updating tool (or possibly make
a text based one) that would change everything for me in the dynamic
realm at least. But trying to maintain a static IP with dynamic
services is kind of a pain.
It doesn't seem to happen like IPv4 when /only/ my router resets or is rebooted.
Deutsche Telekom is funny. For consumer DSL with VoIP IPv6 is
inclusive. If you got the cheapest business DSL with VoIP (technically
the same like the consumer one), you don't get IPv6. And you can't get
it as an option for a few bucks more. If you want IPv6, you have to
get a more expensive business DSL.
It's time that the EU communications regulation enforces that any
internet access product has to provide IPv4 and IPv6.
https://addons.mozilla.org/nl/firefox/addon/ipvfox/
I find it encouraging how many content providers already support
IPv6.
I'm gonna snag that up right now, actually. Sounds like fun. [30
seconds go by]
I dig it. When I hit my own website I get a nice big 6/little 6
display to the right of the address bar. I'll make sure to keep an eye
on it while browsing. Thanks for the link! :)
I am not going to hold my breath either, but I think we will
eventually see the train coming on steam. It is unavoidable.
It's definitely unavoidable at some point. But the wait could continue
to be quite hefty. I hope not, though.
It appears I was wrong. The Ubee modem/router DOES have IPv6
access to its web interface. But... not from a Win Xp system. On
my laptop running Win 7, I can access the web interface via the
link local address,
Via the router's link local address, correct?
Did you have to go to the direct link? For example, when I hit "192.168.1.1" it goes to "http://192.168.1.1/Main_Login.asp" or
something like that, so did you use "[ipv6 address]/Main_Login.asp"?
or did the redirection work properly for you?
Also, did you have to use brackets around the IPv6 address? I remember this being an issue at one point, but some software seems to not
require them.
Not all of the bigger ones and not in all countries. But some of
the big ones ARE moving. In the US, Comcast, Verizon and T-Mobile
ARE moving:
Sure they are moving, but at turtle speed. :)
Is there a difference between how SixXS operated compared to he.net? Meaning, it seems as though he.net makes money on this whole ordeal.
They are in fact a DNS provider as well as a hosting facility.
There are 15 nodes in IPV6.IP6 running an he.net tunnel. If
he.net were to go as well, that would heva a serious impact on
the Fidonet IPv6 vommunity. Fortunately I see no signs of he.net
handing in their dinner pail yet...
I thought he.net started out as a DNS provider and hosting facility.
With the free IPv6 stuff, it lured many people to leave their current providers and go with he.net. So if that's the case, I can't see
he.net going anywhere any time soon either.
Did SixXS do it this way? If not, and it was always a free service
with no income, I can see why the guy got bored with it.
:f1d0:2:221:6 (2:221/6)
I must say that I just love the above. Beautiful!
Have you called your ISP? If so what was their response?
I asked once. The answer was something like "and how would it make
your life better with the ipv6 support in your adsl line?"
Unfuckingbelievable.
Indeed. There is no compettion on the dsl,
so you can take your bussines elsewhere?
My main firewall/router/server has one Gigabit Ethernet port. That
port is connected to a switch that supports VPN's and VPN trunking.
Strange ... in Austria they only offer IPv6 capable CPEs sinceIt would seem then that Austria is ahead of The Netherlands.
some time now. The only way to avoid DS-Lite is to use an older
CPE, which are not offered any more.
Perhaps they have even less IPv4 left?
You may be the first DS-Lite node. ;-)
My main firewall/router/server has one Gigabit Ethernet port. That
port is connected to a switch that supports VPN's and VPN trunking.
VPN or VLAN?
Is there a difference between how SixXS operated compared to he.net?
Meaning, it seems as though he.net makes money on this whole ordeal.
They are in fact a DNS provider as well as a hosting facility.
Did SixXS do it this way? If not, and it was always a free service
with no income, I can see why the guy got bored with it.
The IPv6 may be different. It has not changes since I activited my
present modem/router last week. It is a bit too early to tell how
stable it will be. A guy on the Dutch IPv6 forum who lives near here
and has the same CPE, says his prefix has changed twice in the last
six month.
He too says, the change happened with no obvious link to something he
did.
He also says he has a script to automatically update the DNS. It is
tuned to the API od his DNA provider so it will probably not work for
me. Anyway, if it only changes twice a year, automatic update is not
high on my lisy of priorities.
Indeed. There is no compettion on the dsl,
No. But as stated before, so far I'm happy with the HE tunnel.
so you can take your bussines elsewhere?
My business IS elsewhere. <g>
If the market fails we have to solve the problem by regulation.
Not for data. Based on the latest developments roaming costs for data could even increase.
The mobile providers got good lobbyists and the
politcians forgot about their promises and the EU citizens.
Strange ... in Austria they only offer IPv6 capable CPEs since
some time now. The only way to avoid DS-Lite is to use an older
CPE, which are not offered any more.
It would seem then that Austria is ahead of The Netherlands.
Perhaps they have even less IPv4 left?
I do not think so.
The same company has made their CPE config on xDSL IPv6 ready, but
never activated it on xDSL.
The Telco incumbent does not offer IPv6 for residential at all ...
(they made a IPv6 config 4 years ago, but did not roll it out until
now.
You may be the first DS-Lite node. ;-)
Not if I can evade it ;)
The same company has made their CPE config on xDSL IPv6 ready,Maybe because these "IPv6 ready" CPEs were not as ready as was claimed after all? Here several IPSs have made a false start because they ran
but never activated it on xDSL.
into problems with IPv6 anabled on the CPE. So it was quickly rolled
back.
You may be the first DS-Lite node. ;-)
Not if I can evade it ;)Do they have "zwangsrouter" in Austria?
If so, you may not be in a position to avoid it..
The IPv6 may be different. It has not changed since I activited
my present modem/router last week. It is a bit too early to tell
how stable it will be. A guy on the Dutch IPv6 forum who lives
near here and has the same CPE, says his prefix has changed twice
in the last six month.
I've had it change on me twice in two weeks, and now this time around
it hasn't changed in almost a month.
So I'm unsure as to whether they're still messing around with stuff on their end (most likely), or why it's changing at completely random intervals yet.
He also says he has a script to automatically update the DNS. It
is tuned to the API od his DNA provider so it will probably not
work for me. Anyway, if it only changes twice a year, automatic
update is not high on my lisy of priorities.
It's not aggrivating yet, so I'll stick with what I'm doing currently.
If things get more stable, I may go back to using the fancy static
fido IPv6 address. But as long as my prefix keeps changing at random,
I'd rather just stick with the dynamic setup and have to manually
update less things when the need arises.
so you can take your bussines elsewhere?
My business IS elsewhere. <g>
How about taking your hobby elsewhere? <g>
VLANs or tunnels. But the issue mentioned first, it being on the far
end of a USB bus means it will have issues with high traffic rates, which makes it somewhat unsuitable for being a router on modern networks.
What kind of "high traffic rates" and how "modern" of a network are we talking here? If you're on the business end of things, sure maybe so.
But for a simple household network with capped internet speeds?
How about taking your hobby elsewhere? <g>
Not possible at this geographical location. Well, for a lot of money
it could be possible.
How about taking your hobby elsewhere? <g>
Not possible at this geographical location. Well, for a lot of money
it could be possible.
So the dsl provider has a monopoly. I think a potential competetitor could succesfully claim access to the copper waiving EU rules.
But that isn't going to help you get IPv6 on your dsl isn't it?
Maybe because these "IPv6 ready" CPEs were not as ready as was
claimed after all? Here several IPSs have made a false start
because they ran into problems with IPv6 anabled on the CPE. So
it was quickly rolled back.
First tests did not show any issues.
But it was never rolled out excpet a small trial group, so it may be possible that some issues lurked there.
Do they have "zwangsrouter" in Austria?
If so, you may not be in a position to avoid it..
Yes.
The cable the router is owned by the Telco.
Same with xDSL, but there you get the user/pass to use a different
CPE. (For vectoring enabled VDSL2 lines you need to use the
whitelisted CPE).
In both cases you can use a router behind the CPE.
(Bridge mode in cable and "Singleuser" in xDSL)
I have 5 public IPv4 addresses in the 10/1M ADSL,
he.net tunnel speed is exactly the same than ADSL speed.
I have no obsession about getting native IPv6. :D
What kind of "high traffic rates" and how "modern" of a network are
we talking here? If you're on the business end of things, sure maybe
so. But for a simple household network with capped internet speeds?
I live in .se, so the comes-with-the-housing internet is 100/100,
which for many practical purposes doesn't feel capped at all, and is likely to overwhelm an RPi router.
Especially if you want to do filtering, nat, shaping or other things
to the traffic while passing through and not just bilndly passing
packets over from one port to the next.
First tests did not show any issues.Not one that was published anyway...
First tests did not show any issues.
Not one that was published anyway...
I wrote the configuration for the CPE, and was part of the testing
team ... I would have noticed if there were any issues ;)
I wrote the configuration for the CPE, and was part of theAHA! But then if you were part of the beta team, were you not told why
testing team ... I would have noticed if there were any issues ;)
it was discontinued?
In your case I definitely agree. However, here in the states there are still 3mb/s / 384k/s speed packages available to people that don't want
to pay higher prices for more speed. I would love to have 100/100, but
at the moment the highest my ISP goes is 50/5, so that's what I have.
Agreed there too. However for a hobbiest with slower internet, it's at least possible. I'm not going to try it, though. :)
Michiel van der Vlist wrote to Tommi Koivula <=-
I only got about 50% of the IPv4 speed with my tunnels. Never could figure out why. Now with native IPv6 I have the full speed on IPv6.
I have no obsession about getting native IPv6. :D
You don't have to be obsessed to stand for the cause. :)
AHA! But then if you were part of the beta team, were you not
told why it was discontinued?
Nope.
It was an internal test ... it never was rolled out to real customers. Also all employees got IPv6 training, and than nothing followed ...
I only got about 50% of the IPv4 speed with my tunnels. Never
could figure out why. Now with native IPv6 I have the full speed
on IPv6.
I used to get reasonable speed with tunnels,
but horrible latency (PoP being on the other side of the planet).
There were local tunnel POPs, but they tended to come and go over the years.
I have no obsession about getting native IPv6. :D
You don't have to be obsessed to stand for the cause. :)
I couldn't wait to go native.
It was an internal test ... it never was rolled out to realWhat a waste of time and energy!
customers. Also all employees got IPv6 training, and than nothing
followed ...
The RasPi is a bad choice for a router, since the LAN is actuallyVery good point. The RasPi only has an incoming ethernet port, no outgoing.
USB based.
I have 5 public IPv4 addresses in the 10/1M ADSL,
At no extra cost?
Do you actually use more than 1?
I have no obsession about getting native IPv6. :D
You don't have to be obsessed to stand for the cause. :)
Very good point. The RasPi only has an incoming ethernet port, no
outgoing.
That sentence doesn't really make sense.Maybe not, but I know what I was saying. My current router has an
incoming ethernet port
and 4 outgoing ethernet ports.
The RasPi only has one ethernet port.
It's no problem to run more than one network on top of a singleIf you want to run your network "in nasty ways" that's your choice.
ethernet port, either in nasty ways by just overlaying them
I choose not to.
or with VLANs or tunnels. But the issue mentioned first, it beingWhat kind of "high traffic rates" and how "modern" of a network
on the far end of a USB bus means it will have issues with high
traffic rates, which makes it somewhat unsuitable for being a
router on modern networks.
are we talking here? If you're on the business end of things, sure
maybe so. But for a simple household network
with capped internet speeds?
:f1d0:2:221:6 (2:221/6)I must say that I just love the above. Beautiful!
My main firewall/router/server has one Gigabit Ethernet port. ThatVPN or VLAN?
port is connected to a switch that supports VPN's and VPN trunking.
VPN on a switch? :-)
I have no obsession about getting native IPv6. :D
You don't have to be obsessed to stand for the cause. :)
... he.net certified sage
VPN on a switch? :-)Hey ... those are the new smart switches ;)
Michiel van der Vlist wrote to Tony Langdon <=-
I only got about 50% of the IPv4 speed with my tunnels. Never
could figure out why. Now with native IPv6 I have the full speed
on IPv6.
I used to get reasonable speed with tunnels,
But your IPv4 speed is a "reasonable" 3/1 Mbps isn't it? My IPv4 speed
it 40/4. That is the minmum the cable company has to offer. The high
end is 300/30.
but horrible latency (PoP being on the other side of the planet).
There were local tunnel POPs, but they tended to come and go over the years.
We in Western Europe apparently are in a favorite position. I had a choice of a couple of POPs within a 50km radius...
I have no obsession about getting native IPv6. :D
You don't have to be obsessed to stand for the cause. :)
I couldn't wait to go native.
My account with SixXs is eight years old. So I have been waiting eight years for native IPv6. The tunnels were stable and realiable but natibe is noticeably better. Yes I can imagine why you couldn't not wait
given the less reliable tunnel services in yourpart of the world.
I have no obsession about getting native IPv6. :D
You don't have to be obsessed to stand for the cause. :)
... he.net certified sage
You went through the effort of getting the T-shirt. So you ARE dedicated to
the cause. ;-)
Perhaps I should add some indicator to the list of IPv6 nodes to indicate who's got the T-shirt?
... he.net certified sage
My main firewall/router/server has one Gigabit Ethernet port. That
port is connected to a switch that supports VPN's and VPN trunking.
VPN or VLAN?
VPN on a switch? :-)
:f1d0:2:221:6 (2:221/6)
I must say that I just love the above. Beautiful!
FTS-5004 (http://ftsc.org/docs/fts-5004.001) uses this in all IPv6 examples. And yes, that's the common practice :-)
Not so common as we would like, just 6.5 out of 100
I will add fido6.ddutch.nl as 2001:980:42ea:1:f1d0:2:280:5006
Not so common as we would like, just 6.5 out of 100
Michiels list has only 50 entries. Does that mean there are another 50 where their host name resolves to an IPv6 address?
I will add fido6.ddutch.nl as 2001:980:42ea:1:f1d0:2:280:5006
Good idea.
I will add something like that when I move to vserv.ricsi.priv.at
You went through the effort of getting the T-shirt. So you ARE
dedicated to the cause. ;-)
I just wanted to open the port 25.
The t-shirt was an extra bonus. :)
Good idea.
I will add something like that when I move to vserv.ricsi.priv.at
A hint for nodes running linux and dealing with dynamic prefixes & SLAAC:
ip token set ::2:240:1661:0 dev eth0
A hint for nodes running linux and dealing with dynamic prefixes &
SLAAC:
ip token set ::2:240:1661:0 dev eth0
A hint for nodes running linux and dealing with dynamic prefixes & SLAAC:
ip token set ::2:240:1661:0 dev eth0
ip token set ::2:240:1661:0 dev eth0
Can't you use ::f1d0:2:240:1661 or do you prefer to add the point address?
$ ip token set ::f1d0:1:154:10 dev eth0
RTNETLINK answers: Invalid argument
A hint for nodes running linux and dealing with dynamic prefixes
& SLAAC:
ip token set ::2:240:1661:0 dev eth0
Nice, works like a charm. Only the global scope is affected, the link local remains to be related to the mac address.
Does it stay in effect when you reboot? Or will it have to be added
to an init script or startup file of sorts?
ip token set ::2:240:1661:0 dev eth0
$ ip token set ::f1d0:1:154:10 dev eth0
RTNETLINK answers: Invalid argument
..and yes, my device is indeed eth0. If I can get this to work it
would
make things much easier to maintain, that's for sure.
$ ip token set ::f1d0:1:154:10 dev eth0
RTNETLINK answers: Invalid argument
You need a more recent kernel.
Nice, works like a charm. Only the global scope is affected, the link
local remains to be related to the mac address.
Does it stay in effect when you reboot? Or will it have to be added to an init script or startup file of sorts?
I will add fido6.ddutch.nl as 2001:980:42ea:1:f1d0:2:280:5006
$ ip token set ::f1d0:1:154:10 dev eth0
RTNETLINK answers: Invalid argument
I copied your command string and no error was reported.
Now I see your problen you were not logged in as root as the $ shows
at the start of your example.
..and yes, my device is indeed eth0. If I can get this to work it
would
make things much easier to maintain, that's for sure.
I have put the following in my /etc/network/interfaces and rebooted.
---->8--------
iface eth0 inet dhcp
pre-up ip token set ::f1d0:2:280:5006 dev $IFACE
---->8----------
That will only work for Debian type configuration files.
From 'man ip-token'
From this I gather, that the value is set in the active kernel, not
in the image file. So you will have to set the value with your init scripts.
I will add fido6.ddutch.nl as 2001:980:42ea:1:f1d0:2:280:5006
ip token set ::2:240:1661:0 dev eth0
Can't you use ::f1d0:2:240:1661 or do you prefer to add the point
address?
That was just an example ;)
Well all three DNS servers give the proper answer, so you will have
to wait till the caches between your binkd and de nameservers expire.
I will enable the old address for a few more days.
Well all three DNS servers give the proper answer, so you will have
to wait till the caches between your binkd and de nameservers expire.
I will enable the old address for a few more days.
There is no no magic in adding multiple addresses to an interface in
a Linux system. An early use, was for virtual websites, so every site
had it's own IP address.
Well, I have done it anyway.
There is no no magic in adding multiple addresses to an interface in
a Linux system. An early use, was for virtual websites, so every site
had it's own IP address.
Michiel van der Vlist wrote to Tommi Koivula <=-
Perhaps I should add some indicator to the list of IPv6 nodes to
indicate who's got the T-shirt?
I will add fido6.ddutch.nl as 2001:980:42ea:1:f1d0:2:280:5006
Good idea.A hint for nodes running linux and dealing with dynamic prefixes &
I will add something like that when I move to vserv.ricsi.priv.at
SLAAC:
ip token set ::2:240:1661:0 dev eth0
There is no no magic in adding multiple addresses to an
interface in a Linux system. An early use, was for virtual
websites, so every site had it's own IP address.
Yes ofcourse, dhcp can only provide you with one address, if you want
to add additional ones, they have to be static.
ip token set ::2:240:1661:0 dev eth0
Perhaps I should add some indicator to the list of IPv6 nodes to
indicate who's got the T-shirt?
I'd have to change DNS providers to do that. :(
Perhaps I should add some indicator to the list of IPv6 nodes
to indicate who's got the T-shirt?
I'd have to change DNS providers to do that. :(I don't think so. To participate in the certification project, you
need an account with he.net (that's free) and you need an IPv6
connection. But for the IPv6 connection, you do not need to be
connected through them.
Added a static entry, now I only need to move the Fido system there
... $ ping6 2a03:4000:13:1fd:f1d0:2:310:31 PING 2a03:4000:13:1fd:f1d0:2:310:31(2a03:4000:13:1fd:f1d0:2:310:31) 56 data bytes 64 bytes from 2a03:4000:13:1fd:f1d0:2:310:31: icmp_seq=1 ttl=56 time=40.2 ms
Michiel no reason to try just yet, BinkD is not yet configured or
started there ...
And you need (for the later tests) IPv6 reachable nameservers with
correct glue record.
I assumed Tony was referring to that.
And you need (for the later tests) IPv6 reachable nameservers with
correct glue record.
I assumed Tony was referring to that.
Michiel van der Vlist wrote to Tony Langdon <=-
I don't think so. To participate in the certification project, you need an account with he.net (that's free) and you need an IPv6 connection.
But for the IPv6 connection, you do not need to be connected through them.
Richard Menedetter wrote to Michiel Van Der Vlist <=-
And you need (for the later tests) IPv6 reachable nameservers with
correct glue record.
I assumed Tony was referring to that.
As a matter of fact, as long as you somewhat know what you're doing in regards to IPv6, the entire test could probably be spoofed to a point.
If you didn't have IPv6 at your location, you could just VNC or SSH to another location that does, and run everything from there. I'm not
saying it's the RIGHT thing to do, but it can probably be done.
I know, it's just the sage level of the test I can't get through. :(
I don't think so. To participate in the certification project,
you need an account with he.net (that's free) and you need an
IPv6 connection. But for the IPv6 connection, you do not need to
be connected through them.
I know, it's just the sage level of the test I can't get through. :(
Richard Menedetter wrote to Tony Langdon <=-
What is the problem?
Just slap postfix and apache2 (or nginx) onto any computer (even a
RasPi).
For the DNS test you can also temporarily change to he.net.
That won't do much good unless you find a way to have 5006 answer to
both the old and the new address.
That won't do much good unless you find a way to have 5006 answer
to both the old and the new address.
You can assign more than one address to the same interface. mbcico answers both the SLAAC address and my custom address here.
You can assign more than one address to the same interface. mbcico answers both the SLAAC address and my custom address here.
Yes, one can assign more than one address to an interface. But as I
wrote in another massage, apparently Linux is a bit more flexible than Windows. When I wrote the above I was not aware that in Linux one can
have both the SLAAC address and a fixed address.
A hint for nodes running linux and dealing with dynamic prefixes &
SLAAC:
ip token set ::2:240:1661:0 dev eth0
Sysop: | Nelgin |
---|---|
Location: | Plano, TX |
Users: | 615 |
Nodes: | 10 (1 / 9) |
Uptime: | 53:07:14 |
Calls: | 9,850 |
Calls today: | 5 |
Files: | 96,973 |
Messages: | 1,080,728 |
Posted today: | 1 |