Unless you are running as an SLA, I doubt it.

A /64 in IPv6 is a prefix length. A single IPv6 address is 128bits in length. So /64 is half of that. Commonly end sites or host on the LAN are handed /64, and the remaining half is taken from the host MAC address. Joining them together get's you a single IPv6 address (128bits).

This is where SLAAC comes in, as it autoconfigures the host address by combining the /64 prefix length and host MAC address. No user intervention required, hence Stateless Auto Address Configuration (SLAAC).

You can also think of a /64 as a /24 class C in IPv4-land.

Hope this helps :>

This post has been edited by rogue: Nov 28 2013, 07:03 PM

So complicated..see also pening lol...that is meant for network security ppl or ppl involve in network design software engineers..



This 1 abit simpler rite ?


This post has been edited by rogue: Nov 28 2013, 10:01 PM

Btw, today I checked and I think my area(Subang Jaya) is covered already. Tried yesterday but nothing.

I disconnected/reconnected my connection and I can see v6 addresses.

Sounda like good news for gamers..

So in other words, a /64 is the norm that everyone gets, and /48, /56 or /60 are for 'special' cases/scenarios..

Hi,

I would first try connecting a laptop/pc directly to your BTU and dial PPPoE and see if you can get an IPv6 address just to confirm IPv6 is supported in your area.

If you have it, then you could try this in PfSense:



Quote above taken from this link.

The link above is base on Comcast IPv6 setup but should give you a rough idea for your setup. One common problem ppl having in that forum is that they get IPv6 on both WAN/LAN, but no IPv6 default route(IPv6 gateway). So if this happens to you, maybe you need to script your way into adding an IPv6 default route after WAN is established.

Hope this helps..

Hello,

For MTU, I'm using OpenWRT Backfire and it's currently just using the default MTU 1492 on PPPoE iface with no disconnects or stability issues so far.

For stability I fired up XChat and connect to 2 different IPv6 IRC networks without joining any channels. Just keeping the status window open to see if either networks disconnect. So far their still connected. I use IRC as a test to see how stable TM IPv6 is coz IRC is real time..



Or maybe you can lower your lease time to something between 600-900..that way it will expire the old lease quickly automatically. I have mine set to 1800 in dhcp6s and radvd, both preferred and max lifetime.

Ofcourse this is only during testing periods. Once your IPv6 is stable enough, you should increase back the lease time.

For the lease setting, there should be something like in IPv4 dhcp for you to specify the lease time. Your router's default could be somewhere between 3600 - 7200 or 1 day.


I get the same thing too. It assign's IPv6 to the WAN interface roughly 5-10mins later. I think it has to do with the current dhcpv6 lease time on the WAN side. If your connection disconnects before the lease expires, then it has to wait for that lease to expire before handling out a new lease.

Yeah. The lease will be the same as long as you don't disconnect.

You could try confirming with TM what DHCPv6 lease lifetime are they using so we can better understand what happens when it happens.

This post has been edited by rogue: Dec 3 2013, 04:22 PM

Hi,

If u using ipv6-test.com sometimes it will show no IPv6 eventhough you have it. For this site the trick is to use only IPv6 dns and temporarily disable IPv4 dns before running the test. Enable back after the test.

Other sites like test-ipv6.com or ipv6test.google.com don't have this symptom I notice. You might want to run test on these 2 sites first to confirm working IPv6 before running at ipv6-test.com

Hi, if you go into OpenWRT -> Network -> Interfaces, does your br-lan interface have a 2001:e68 address ?

If you don't have the 2001 address on either LAN or WAN after configuring IPv6 it could mean your area have no IPv6 yet.

Hi,

Where does your IPv6 default route point to ?

When you say:



What do you mean ? If your IPv6 traffic comes from TM and you want it to go through your ssh/openvpn tunnel instead, this won't work unless you have source routing enabled. I don't see why you want to do this..

Please clarify..

This post has been edited by rogue: Dec 9 2013, 11:33 AM

Hi,

Ok understand now. I assume you are using socks5 with ssh tunneling. Only v4 traffic will go through the tunnel. v6 traffic will use the default ipv6 route instead. If I understand correctly you need some sort of translation mechanism inbetween because if v6 traffic goes through the tunnel but the final destination is only v4, there is no way to translate it back to you since the origin is v6.

If the destination is dual-stack and reachable, then it will use v6 instead via your ipv6 default route.

Hi nymerine,

I just tried your setup and it seems to work. My bad. I always thought socks5 worked on v4 alone.

I tried it on a terminal and Firefox:

ssh -6 -D 1011 user@remote_host

Then I confirm it with 'netstat -tunpl | grep 1011' and I see the port bound to 127.0.0.1:1011

Then in firefox, under socks5 I just set socks host to localhost and port to 1011

www.whatismyip.com shows the tunnel v4 address and www.whatismyipv6.com shows the tunnel v6 address.

Try downloading Firefox from here

I currently get full speed on VIP10

Does it work in Safe Mode ? Could be a windows bug or a driver issue or an antivirus..

I'm running OpenWRT Backfire on WR1043ND and don't have these symptoms. I use some scripting as a workaround. I've disabled getting IPv6 address on the WAN interface. I only want PD for LAN.

I've also created a custom script and use that with /etc/hotplug.d/ so it executes the script when the interface re-establishes. It will update my dhcp6s.conf and restart dhcp6s. For radvd you can have a fix config. Just replace the prefix field with '::/64' so it will substitute with whatever PD /64 you get on the LAN side.

Both dhcp6s and radvd have low lease time, so the client gets a new lease as soon as possible.

I also added a second ipv6 default route for redundant purposes and the route is 2000::/3 because I notice sometimes the current default route(remote link-local address fe80::/10) goes missing or removed whenever connection is re-established. So adding another route entry 2000::/3 as backup helps keep my connection active.

Hi..you have v6 on both WAN and LAN interface ?

I don't have this problem in my setup. I only request dhcp-pd which I get a /64 and my WAN has no v6 address. My WAN MTU is 1492 and LAN is 1500. I'm using OpenWRT if it matters.

No stuttering/buffering issues when accessing Youtube.

Hi..I'm still on Backfire btw..

I've no GUI setup..my setup is scripted..so I dont have issues getting dynamic dhcp-pd as my scripts take care of flushing/reassigning the new /64 prefix if I reconnect.

I'm using radvd together with Wide dhcp client/server btw

I also have 2 default routes added for redundancy as I notice the current default route seems to dissappear after awhile when the lifetime of the dhcp lease expires.