They had a list of what prefixes were located in what region back when they whitelisted dns servers. HE should be able to provide that again to Google if Google can do anything with it for the not-whitelisted stuff. I used to just email Google about broker user complaints and a few hours/days later it would be resolved. Pretty sure someone at HE can do that too. Can supply email if it isn't obvious to them.

chi = chicago

This a 2wire "modem"? There are threads about those devices here, maybe they'll help if it is.

https://plus.google.com/u/0/116899029375914044550/posts/J56HMSZ316t

And there is also http://www.google.com/ncr

anything in iptables? (not ip6tables)

So your public ipv4 address is configured on your linux box? Or more than likely, is it actually behind NAT and you didn't change the source IP to the NAT IP when pasting the tunnel commands

Before it wasn't reachable. now it is!

Code:

HOST: ipvsixme                    Loss%   Snt   Last   Avg  Best  Wrst StDev
  1.|-- 2001:470:1:9::1            0.0%     2    0.6   0.6   0.6   0.6   0.0
  2.|-- 10gigabitethernet8-4.core  0.0%     2    0.6   0.6   0.6   0.6   0.0
  3.|-- 10gigabitethernet1-1.core  0.0%     2    1.0   1.0   1.0   1.0   0.0
  4.|-- 10gigabitethernet3-3.core  0.0%     2   28.0  28.0  28.0  28.0   0.0
  5.|-- 10gigabitethernet5-5.core  0.0%     2   39.6  39.6  39.6  39.6   0.0
  6.|-- tserv1.mci3.he.net         0.0%     2   40.0  40.1  40.0  40.3   0.2
  7.|-- 2001:470:7c:12e::1         0.0%     2   99.5  86.0  72.5  99.5  19.1

That solve the issue for your clients sitting on the LAN?

Ok based on testing 2001:470:7b:12e::1, I see that what you are using is indeed your ROUTED prefix. Make certain you've configured an address out of the /64 on eth0, like 2001:470:7c:12e::1/64

Did you use the ROUTED prefix for the LAN client machines? Paste some configs.

I found slightly different PF rules from 2011 (google cache), and also not certain why the protocol 43 since IPv6 anything would be encapsulated in p41 packets.

http://webcache.googleusercontent.com/search?q=cache:GOjXWRitME4J:https://johan.pp.se/2011/06/06/freebsd-pf-and-ipv6+&cd=2&hl=en&ct=clnk&gl=us

Yay obfuscated IPs, that is always really helpful.

The *.*.*.225, is that your real-world IP? Is the machine actually configured with it?
Can you ping6 HE's side of the tunnel?
I don't know PF's syntax, but where is the rule that allows Protocol 41? And I mean specifically Protocol 41 traffic, which is IPv6 encapsulated in IPv4.

pmtud issue? what does a tracepath6 look like for output?

http://bgp.he.net/AS51377

Single homed behind Cogent. The same Cogent who doesn't peer with HE or buy transit from anyone Enjoy the limited IPv6 connectivity!

Looks like they even took down all their peering/transits, going down from 16 to 1.

need an ipv6 address on the vboxnet0 interface out of that routed /64

i appreciate your help but this has gotten too complicated. How can i direct a domain to point to another hosting site?

If you are just going to have someone do all the legwork and get stuff working, then yes: 'what is the point? Do I need to do this?' makes it irrelevant almost.

If you are doing this to get IRC/SMTP access, then the point is to learn some internet service hosting fundamentals. If you can get it working with IPv4, then it should be relatively similar/easy for IPv6. You could try reading some Microsoft Press books/documentation for IIS and go from there.