written for, and discussion at, http://www.tunnelbroker.net/forums/index.php?topic=2293.0
It would appear that AT&T has quietly added IPv6 support through 6rd for, seemingly, ALL of its customers without telling anyone (specifically, without telling any of the said customers).
All existing customers “are not affected” and don’t know about any such support, since it’s not actually supported by most CPE.
Details are here:
http://www.dslreports.com/forum/r26841639-
If you have Motorola NVG510, it’s enabled by default (yes, that’s how we, the customers, found out!). If you have 2Wire PoS, you can still play around and have it configured through your own equipment (and your equipment doesn’t even has to have 6rd support, e.g. OS X and OpenBSD would do just fine for basic 6rd IPv6 connectivity).
Basically, [b]2602:300::/28[/b] (6rdPrefix/6rdPrefixLen) and [b]12.83.49.81[/b] (6rdBRIPv4Address, which is an anycast) is all you need to get it running, IPv4MaskLen is 0 (use the whole IPv4 address within IPv6, but notice that due to 6rdPrefixLen being /28 (instead of the more conventional /32) you have to do some one-nibble shifting, but the plus side is that you do get a /60 in the end). Plus [b]74.82.42.42[/b] (ordns.he.net.), of course. :-) You can also try the standard dnsr{1,2}.sbcglobal.net, but they have a few problems: http://www.dslreports.com/forum/r26902814-IPv6-6rd-DNS-Cannot-resolve-IPv6-only-zones.
This 6rd from AT&T works great in the Bay Area, but I’ve heard that IPv4-wise, Los Angeles gets routed through San Jose, which is a bummer, since IPv6-wise, HE.net is routed through Los Angeles from San Jose AT&T, hence it would appear like people in LA would have to make three full roundtrips (LA -> SJ over IPv4 (1 round), then SJ -> LA -> SJ over IPv6 (2 rounds)) to reach HE’s FMT IPv6 resources over this 6rd. :-)
[b]This is a must-try if you’re on AT&T![/b]
Here’s a collection of my traceroute6’s:
http://www.dslreports.com/forum/r26907835-IPv6-latency-BayArea-Local-IPv6-around-San-Jose-
Just in case you don’t care for external links:
[code]
% traceroute6 www.google.com
Wed 22 Feb 2012 18:38:09 PST
traceroute6 to www.l.google.com (2001:4860:4001:801::1011) from 2602:306:37cY:YYY0::1, 30 hops max, 12 byte packets
1 2602:300:c533:1510::5 (2602:300:c533:1510::5) 1.579 ms 1.368 ms 1.309 ms
2 sj2ca404me3.ip.att.net (2001:1890:ff:ffff:12:122:119:192) 4.014 ms 4.042 ms 4.007 ms
3 2001:1890:c00:8c02::1116:e9cb (2001:1890:c00:8c02::1116:e9cb) 54.538 ms 55.262 ms 55.846 ms
4 2001:4860::1:0:21 (2001:4860::1:0:21) 5.572 ms 5.374 ms 5.154 ms
5 2001:4860:0:1::1af (2001:4860:0:1::1af) 5.425 ms 5.55 ms 5.343 ms
6 2001:4860:8000:e:92e6:baff:fe53:a202 (2001:4860:8000:e:92e6:baff:fe53:a202) 56.341 ms 57.198 ms 56.167 ms
[/code]
Here’s my traceroute / mtr to the 6rdBRIPv4Address:
[code]
% traceroute -I 12.83.49.81
Wed 22 Feb 2012 18:47:16 PST
traceroute to 12.83.49.81 (12.83.49.81), 32 hops max, 60 byte packets
5 12.83.39.137 (12.83.39.137) 3.581 ms 1.700 ms 1.568 ms
6 12.83.49.81 (12.83.49.81) 1.311 ms 1.310 ms 1.107 ms
% mtr -c60 —report{,-wide} 12.83.49.81
Wed 22 Feb 2012 18:49:30 PST
HOST: 99-124-xxx-xxx.uvs.sntcca.sbcglobal.net Loss% Snt Last Avg Best Wrst StDev
1.|— 99-124-xxx-xxx.uvs.sntcca.sbcglobal.net 0.0% 60 0.8 0.4 0.3 0.8 0.1
2.|— 76-220-32-3.lightspeed.sntcca.sbcglobal.net 76.7% 60 2.2 2.5 1.9 5.1 0.9
3.|— 71.145.0.104 98.3% 60 3.3 3.3 3.3 3.3 0.0
4.|— 71.145.0.80 96.7% 60 6.7 6.8 6.7 6.9 0.1
5.|— 12.83.39.137 0.0% 60 1.7 2.1 1.7 2.8 0.2
6.|— 12.83.49.81 0.0% 60 1.6 1.4 1.2 1.6 0.1
[/code]
Here’s traceroute6 to he.net and back (as I mentioned, it goes SJC -> LAX -> FMT, since LA is the closest point where HE and AT&T do IPv6 peering):
[code]
% traceroute6 ns2.linode.com
Wed 22 Feb 2012 19:07:19 PST
traceroute6 to ns2.linode.com (2600:3c01::a) from 2602:306:37cY:YYY0::1, 30 hops max, 12 byte packets
1 2602:300:c533:1510::5 (2602:300:c533:1510::5) 1.579 ms 1.2 ms 1.315 ms
2 la2ca02jt.ip.att.net (2001:1890:ff:ffff:12:122:127:43) 129.526 ms 13.831 ms 24.7 ms
3 10gigabitethernet5-2.core1.lax2.he.net (2001:470:0:1e6::1) 13.755 ms 13.938 ms 13.77 ms
4 10gigabitethernet2-1.core1.lax1.he.net (2001:470:0:72::1) 13.843 ms 13.829 ms 13.787 ms
5 10gigabitethernet7-4.core1.fmt2.he.net (2001:470:0:18d::1) 24.908 ms 21.798 ms 21.9 ms
6 gige-g4-18.core1.fmt1.he.net (2001:470:0:2d::1) 22.01 ms 21.99 ms
linode-llc.10gigabitethernet2-3.core1.fmt1.he.net (2001:470:1:1db::2) 22.852 ms
7 ns2.linode.com (2600:3c01::a) 22.439 ms 22.319 ms 22.335 ms
[/code]
[code]
# traceroute6 2602:306:37cY:YYY0::1
traceroute to 2602:306:37cY:YYY0::1 (2602:306:37cY:YYY0::1), 16 hops max, 80 byte packets
2 10gigabitethernet2-3.core1.fmt1.he.net (2001:470:1:1db::1) [AS6939] 5.666 ms 5.782 ms 5.760 ms
3 gige-g4-8.core1.fmt2.he.net (2001:470:0:2d::2) [AS6939] 0.454 ms 0.443 ms 0.532 ms
4 10gigabitethernet6-4.core1.lax1.he.net (2001:470:0:18d::2) [AS6939] 8.514 ms 8.500 ms 8.474 ms
5 10gigabitethernet1-3.core1.lax2.he.net (2001:470:0:72::2) [AS6939] 15.265 ms 8.905 ms 14.903 ms
6 att-internet4-as7018.10gigabitethernet5-2.core1.lax2.he.net (2001:470:0:1e6::2) [AS6939] 8.915 ms 8.988 ms 9.052 ms
7 * * *
8 * * *
9 2001:1890:ff:ffff:12:122:114:41 (2001:1890:ff:ffff:12:122:114:41) [AS7018] 22.141 ms 22.114 ms 22.078 ms
10 2602:300:c533:1510::5 (2602:300:c533:1510::5) [*] 21.171 ms 21.174 ms 21.148 ms
11 * * *
12 * * *
13 * * *
14 2602:300:c533:1510::6 (2602:300:c533:1510::6) [*] 22.154 ms 22.425 ms 22.649 ms
15 2602:300:c533:1510::5 (2602:300:c533:1510::5) [*] 22.280 ms 22.709 ms 23.197 ms
16 2602:306:37cY:YYY0::1 (2602:306:37cY:YYY0::1) [*] 22.543 ms 22.710 ms 23.209 ms
[/code]
Try the following to get your IPv6 prefix from the IPv4 address:
[code]
% printf “%02x%02x%02x%02x” 99 124 xxx xxx | awk ‘{print “2602:30” substr($1,1,1) “:” substr($1,2,4) “:” substr($1,6) “0::/60”}’
2602:306:37cY:YYY0::/60
[/code]
What would be interesting to know are the locations of where 6rdBR’s have been deployed. So far, it’s known that LA uses the one in San Jose, so there doesn’t appear to be a 6rdBR in LA. You can guesstimate the location of your 6rdBR by doing `traceroute 12.83.49.81` from your local AT&T network, or by doing a traceroute6 of your 6rd IPv6 address from a remote network, seeing the latency of the last hop of the 2602:300:c533:1510::/60 origin (which is, you’ve guessed it, the anycast 6rd prefix of the 6rdBR’s 12.83.49.81). Your total latency for a given IPv6 resource over this 6rd would be the sum of these two latencies around the 6rdBR anycast (12.83.49.81 / 2602:300:c533:1510::/60).
Happy 6rd tunnelling!
P.S. I should warn you that this whole info is based solely on user-submitted “reverse-engineered” information. I should also warn you that if you ever decide to go to AT&T’s corporate IPv6 web-site (the link for which is very short and memorable, but is intentionally omitted from this post such as to not waste people’s time), that you will not find a single useful piece of information about anything whatsoever, and the time that you will lose you’ll never get back! (-:
