Is there something wrong in the IPv6 network?

Advanced feature discussion, beta programs and unsupported "Labs" features.
4 posts Page 1 of 1
by nsmiller » Sat Oct 22, 2016 8:38 pm
I have a 6rd tunnel. I usually pass test-ipv6.com 10/10. Nevertheless, when accessing ssa.gov, I have to disable IPv6 to use the site; else the site chokes on a login attempt. And when I try to access mail.yahoo.com, Opera, SeaMonkey, and IE12 Spin and spin, then fail. Opera throws this message:

This site can’t be reached
mail.yahoo.com took too long to respond.

Yet disable IPv6, and I have no problem accessing, or logging in.

Next idea is to test the 6in4 tunnel, and to resurrect my Hurricane Electric 6in4 tunnel to see of one of those works.
by wwwdrich » Tue Nov 15, 2016 9:02 pm
I have been seeing IPv6 issues lately as well. Currently I can't get to yahoo (ok, so maybe that's a feature :-) ) over IPv6 from my home network, but other sites work fine. I'm seeing very odd things with traceroute as well. As an example, I can ping my server at home from an external host just fine, but a traceroute to that same system fails on one of Sonic's devices.

Code: Select all

> ping6 2602:243:2002:7f30:4000::20 
PING6(56=40+8+8 bytes) 2001:1868:a000:17::142 --> 2602:243:2002:7f30:4000::20
16 bytes from 2602:243:2002:7f30:4000::20, icmp_seq=0 hlim=56 time=28.133 ms
16 bytes from 2602:243:2002:7f30:4000::20, icmp_seq=1 hlim=56 time=27.709 ms
16 bytes from 2602:243:2002:7f30:4000::20, icmp_seq=2 hlim=56 time=20.611 ms
16 bytes from 2602:243:2002:7f30:4000::20, icmp_seq=3 hlim=56 time=24.598 ms
16 bytes from 2602:243:2002:7f30:4000::20, icmp_seq=4 hlim=56 time=32.279 ms
^C
--- 2602:243:2002:7f30:4000::20 ping6 statistics ---
5 packets transmitted, 5 packets received, 0.0% packet loss
round-trip min/avg/max/std-dev = 20.611/26.666/32.279/3.891 ms

> traceroute6 2602:243:2002:7f30:4000::20
traceroute6 to 2602:243:2002:7f30:4000::20 (2602:243:2002:7f30:4000::20) from 2001:1868:a000:17::142, 64 hops max, 12 byte packets
 1  2001:1868:a000:17::1  0.621 ms  6.319 ms  0.624 ms
 2  xe0-0-0-4.core3.scl.layer42.net  1.041 ms  0.587 ms  0.542 ms
 3  xe0-0-0-0-100.core1.sv8.layer42.net  1.267 ms  0.941 ms  0.901 ms
 4  2001:1868:a200::2  2.428 ms  1.888 ms  1.756 ms
 5  0.xe-6-0-0.gw.equinix-sj.sonic.net  1.341 ms  1.099 ms  1.081 ms
 6  201.ge-1-2-0.cgn-gw.equinix-sj.sonic.net  1.191 ms  1.390 ms  1.160 ms
 7  * * *
 8  * * *
^C
by wwwdrich » Thu Jan 19, 2017 6:07 pm
I really wish Sonic provided better IPv6 support -- this post has been here since Oct with no replies and the last time I checked we can't open tickets for IPv6 issues.

I'm still seeing issues with yahoo after all this time. I'm guessing it's a problem with the routes that they are advertising, but it would be nice to get some confirmation from Sonic so I can take the issue to yahoo!

Code: Select all

> traceroute6 mail.yahoo.com
traceroute6 to fo-ds-ats.member.g02.yahoodns.net (2001:4998:c:e33::50) from 2602:243:2002:7f30::, 64 hops max, 12 byte packets
 1  * *^C

>  traceroute6 yahoo.com
traceroute6: Warning: yahoo.com has multiple addresses; using 2001:4998:44:204::a7
traceroute6 to yahoo.com (2001:4998:44:204::a7) from 2602:243:2002:7f30::, 64 hops max, 12 byte packets
 1  cust-gw.ipv6.sonic.net  40.530 ms * *
 2  201.ge-6-1-1.gw.equinix-sj.sonic.net  83.753 ms  94.119 ms  90.779 ms
 3  100gigabitethernet2-3.core1.sjc2.he.net  101.477 ms  104.287 ms  116.266 ms
 4  PAT1.SJC.yahoo.com  115.221 ms
    as10310.bay-area.megaport.com  25.809 ms
    PAT1.SJC.yahoo.com  22.474 ms
 5  ae-4.pat1.dnx.yahoo.com  48.425 ms
    ae-9.pat2.dnx.yahoo.com  58.281 ms  48.764 ms
 6  ae-6.pat2.nez.yahoo.com  57.028 ms
    ae-0.pat1.nez.yahoo.com  58.066 ms
    ae-5.pat2.nez.yahoo.com  58.060 ms
 7  2001:4998:f000:11::1  62.941 ms
    2001:4998:f000:209::1  75.910 ms
    2001:4998:f000:208::1  81.080 ms
 8  2001:4998:44:fe1c::1  58.413 ms
    2001:4998:44:fe1a::1  65.405 ms
    2001:4998:44:fc1a::1  68.890 ms
 9  v6.e18-25.fab3-1-gdc.ne1.yahoo.com  76.177 ms
    v6.e17-1-fab2-1-gdc.ne1.yahoo.com  67.986 ms
    v6.e17-1.fab6-1-gdc.ne1.yahoo.com  65.556 ms
10  2001:4998:44:d403::1  57.344 ms
    2001:4998:44:d202::1  64.349 ms  62.136 ms
11  ir1.fp.vip.ne1.yahoo.com  55.905 ms  67.020 ms  55.593 ms
by tomoc » Fri Jan 20, 2017 9:25 am
Apologies for the late reply, I have at least a couple answers for you.
nsmiller wrote:I have a 6rd tunnel. I usually pass test-ipv6.com 10/10. Nevertheless, when accessing ssa.gov, I have to disable IPv6 to use the site; else the site chokes on a login attempt.
From my initial research, this looks to be an IPv6 issue well outside our network. I picked a random only IPv6 traceroute utility and am having the same reachability issues. (I used http://www.subnetonline.com/pages/ipv6- ... eroute.php to try and trace to ssa.gov)
nsmiller wrote: And when I try to access mail.yahoo.com, Opera, SeaMonkey, and IE12 Spin and spin, then fail. Opera throws this message:

This site can’t be reached
mail.yahoo.com took too long to respond.

Yet disable IPv6, and I have no problem accessing, or logging in.
I don't see a problem with this from a traceroute standpoint, so I'll have to take a look and see what I can find in our lab with a 6rd connection-- I'll get back to you
nsmiller wrote:Next idea is to test the 6in4 tunnel, and to resurrect my Hurricane Electric 6in4 tunnel to see of one of those works.
I'm curious if this works, but fixing 6rd needs to happen anyway
wwwdrich wrote:I'm seeing very odd things with traceroute as well. As an example, I can ping my server at home from an external host just fine, but a traceroute to that same system fails on one of Sonic's devices.

Code: Select all

> ping6 2602:243:2002:7f30:4000::20 
PING6(56=40+8+8 bytes) 2001:1868:a000:17::142 --> 2602:243:2002:7f30:4000::20
16 bytes from 2602:243:2002:7f30:4000::20, icmp_seq=0 hlim=56 time=28.133 ms
16 bytes from 2602:243:2002:7f30:4000::20, icmp_seq=1 hlim=56 time=27.709 ms
16 bytes from 2602:243:2002:7f30:4000::20, icmp_seq=2 hlim=56 time=20.611 ms
16 bytes from 2602:243:2002:7f30:4000::20, icmp_seq=3 hlim=56 time=24.598 ms
16 bytes from 2602:243:2002:7f30:4000::20, icmp_seq=4 hlim=56 time=32.279 ms
^C
--- 2602:243:2002:7f30:4000::20 ping6 statistics ---
5 packets transmitted, 5 packets received, 0.0% packet loss
round-trip min/avg/max/std-dev = 20.611/26.666/32.279/3.891 ms

> traceroute6 2602:243:2002:7f30:4000::20
traceroute6 to 2602:243:2002:7f30:4000::20 (2602:243:2002:7f30:4000::20) from 2001:1868:a000:17::142, 64 hops max, 12 byte packets
 1  2001:1868:a000:17::1  0.621 ms  6.319 ms  0.624 ms
 2  xe0-0-0-4.core3.scl.layer42.net  1.041 ms  0.587 ms  0.542 ms
 3  xe0-0-0-0-100.core1.sv8.layer42.net  1.267 ms  0.941 ms  0.901 ms
 4  2001:1868:a200::2  2.428 ms  1.888 ms  1.756 ms
 5  0.xe-6-0-0.gw.equinix-sj.sonic.net  1.341 ms  1.099 ms  1.081 ms
 6  201.ge-1-2-0.cgn-gw.equinix-sj.sonic.net  1.191 ms  1.390 ms  1.160 ms
 7  * * *
 8  * * *
^C
The last hop you are seeing in that traceroute is the IPv6 endpoint where your traffic gets re-encapsulated and hauled over our IPv4 network. I confess I am not sure whether your 6rd endpoint (modem/router) should be responding to traceroute traffic, but I would not be the least bit surprised if not.
Tomoc
Sonic NOC
4 posts Page 1 of 1

Who is online

In total there are 25 users online :: 0 registered, 0 hidden and 25 guests (based on users active over the past 5 minutes)
Most users ever online was 999 on Mon May 10, 2021 1:02 am

Users browsing this forum: No registered users and 25 guests