[outages] IPv6 tunnels in FRA1 on HE.net down?

Constantine A. Murenin mureninc at gmail.com
Tue May 14 12:42:54 EDT 2013


This just happened again:  all IPv6 tunnels on tserv1.fra1.he.net.
were inaccessible; from within a tunnel, cannot access any IPv6
resource, other than ordns.he.net, which runs on the tunnel server
itself.

Why does FRA1 loses IPv6 connectivity so often?

Update: Seems like it has been resolved as I've been writing this
email, but this would seem to happen a few times too many.

C.

On 23 April 2013 19:28, Constantine A. Murenin <mureninc at gmail.com> wrote:
> On 23 April 2013 18:28, Constantine A. Murenin <mureninc at gmail.com> wrote:
>> As of a couple of minutes ago, my IPv6 tunnel seems to have no
>> connectivity, weirdly other than being able to access ordns.he.net
>> (2001:470:20::2) just fine, but not ns{2,3,4,5}.he.net, or any other
>> IPv6 host.
>
> Update: reported to he.net support@ on 18:34, including a follow-up phone call;
> everything's back online, as of at least 18:51 PT.
> Pretty fast resolution, for a free service. :-)
>
> According to he.net, tserv wasn't responding on its IPv6 address, and has
> henceforth been rebooted.
>
> Which adds up as per my mtr from a Linode:
>
> # mtr --report{,-wide,-cycles=60} --order "SRL BGAWV" -6 XXXXXXXXX ; date
>   2. 10gigabitethernet2-3.core1.fmt1.he.net     60    60  0.0%    0.4   2.5   8.5  85.5  15.2
>   3. 10gigabitethernet1-2.core1.sjc2.he.net     60    60  0.0%    0.8   2.5   5.3  51.6   8.5
>   4. 10gigabitethernet3-3.core1.den1.he.net     60    60  0.0%   27.8  31.7  32.3  70.0   7.3
>   5. 10gigabitethernet5-5.core1.mci3.he.net     60    60  0.0%   39.7  43.6  44.3 114.3  10.1
>   6. 10gigabitethernet5-2.core1.chi1.he.net     60    60  0.0%   52.0  56.2  57.4 177.2  17.0
>   7. 100gigabitethernet7-2.core1.nyc4.he.net    60    59  1.7%   69.1  72.2  72.5 119.8   7.1
>   8. 10gigabitethernet1-2.core1.lon1.he.net     60    60  0.0%  137.8 141.8 142.2 229.2  12.0
>   9. 10gigabitethernet4-2.core1.fra1.he.net     60    60  0.0%  149.4 154.0 154.4 242.8  12.7
>  10. ???                                        60     0 100.0    0.0   0.0   0.0   0.0   0.0
> Tue Apr 23 18:14:29 PDT 2013
>
> ...
>
>   2. 10gigabitethernet2-3.core1.fmt1.he.net     60    60  0.0%    0.4   2.8 9.3 78.3 16.8
>   3. 10gigabitethernet1-2.core1.sjc2.he.net     60    60  0.0%    0.8   2.3   3.9  13.9   3.8
>   4. 10gigabitethernet3-3.core1.den1.he.net     60    60  0.0%   27.8  30.3  30.5  39.2   3.5
>   5. 10gigabitethernet5-5.core1.mci3.he.net     60    60  0.0%   39.7  43.4  43.6  52.8   4.4
>   6. 10gigabitethernet5-2.core1.chi1.he.net     60    60  0.0%   52.0  54.1  54.2  62.2   3.2
>   7. 100gigabitethernet7-2.core1.nyc4.he.net    60    60  0.0%   69.1  71.5  71.6  82.1   3.5
>   8. 10gigabitethernet1-2.core1.lon1.he.net     60    60  0.0%  137.8 140.3 140.4 149.1   3.5
>   9. 10gigabitethernet4-2.core1.fra1.he.net     60    60  0.0%  149.4 152.3 152.3 162.5   3.7
>  10. tserv1.fra1.he.net                         60    60  0.0%  154.1 155.5 155.5 163.4   2.2
>  11. IPv6.XXXXXX                                60    59  1.7%  155.2 156.0 156.1 162.9   1.2
> Tue Apr 23 18:55:34 PDT 2013
>
>
>
> And I guess ordns.he.net (2001:470:20::2) really runs on tserv
> (and hence wasn't affected during the outage).
>
> Cns# echo {ordns,ns{2,3,4,5}}.he.net | xargs -n1 traceroute6 -l; date
> traceroute6 to ordns.he.net (2001:470:20::2) from 2001:470:XXXX:YYYY::2, 64 hops max, 12 byte packets
>  1  ordns.he.net (2001:470:20::2)  6.39 ms  6.495 ms  6.213 ms
> traceroute6 to ns2.he.net (2001:470:200::2) from 2001:470:XXXX:YYYY::2, 64 hops max, 12 byte packets
>  1  XXXXXX.tunnel.tserv6.fra1.ipv6.he.net (2001:470:XXXX:YYYY::1)  9.353 ms  8.783 ms  9.328 ms
>  2  gige-g2-20.core1.fra1.he.net (2001:470:0:69::1)  13.12 ms  15.134 ms  6.252 ms
>  3  10gigabitethernet5-3.core1.lon1.he.net (2001:470:0:1d2::1)  18.113 ms  23.554 ms  20.342 ms
>  4  ns2.he.net (2001:470:200::2)  20.449 ms  22.873 ms  20.517 ms
> traceroute6 to ns3.he.net (2001:470:300::2) from 2001:470:XXXX:YYYY::2, 64 hops max, 12 byte packets
>  1  * XXXXXX.tunnel.tserv6.fra1.ipv6.he.net (2001:470:XXXX:YYYY::1)  8.743 ms  9.2 ms
>  2  gige-g2-20.core1.fra1.he.net (2001:470:0:69::1)  5.934 ms  10.749 ms  6.197 ms
>  3  10gigabitethernet5-3.core1.ams1.he.net (2001:470:0:225::1)  18.567 ms  17.728 ms  13.282 ms
>  4  ns3.he.net (2001:470:300::2)  13.462 ms  13.412 ms  13.525 ms
> traceroute6 to ns4.he.net (2001:470:400::2) from 2001:470:XXXX:YYYY::2, 64 hops max, 12 byte packets
>  1  XXXXXX.tunnel.tserv6.fra1.ipv6.he.net (2001:470:XXXX:YYYY::1)  8.838 ms  8.684 ms  8.682 ms
>  2  gige-g2-20.core1.fra1.he.net (2001:470:0:69::1)  6.233 ms  13.208 ms  5.97 ms
>  3  ns4.he.net (2001:470:400::2)  6.145 ms  6.38 ms  6.384 ms
> traceroute6 to ns5.he.net (2001:470:500::2) from 2001:470:XXXX:YYYY::2, 64 hops max, 12 byte packets
>  1  * XXXXXX.tunnel.tserv6.fra1.ipv6.he.net (2001:470:XXXX:YYYY::1)  8.655 ms  8.848 ms
>  2  gige-g2-20.core1.fra1.he.net (2001:470:0:69::1)  12.607 ms  6.527 ms  11.929 ms
>  3  10gigabitethernet5-3.core1.ams1.he.net (2001:470:0:225::1)  13.289 ms  13.968 ms  16.754 ms
>  4  ns5.he.net (2001:470:500::2)  14.111 ms  13.575 ms  13.385 ms
> Tue Apr 23 18:59:48 PDT 2013
>
>
> However, it's unclear why IPv6 connectivity of tserv1.fra1.he.net
> doesn't seem to be monitored otherwise. :/
>
> Cheers,
> Constantine.



More information about the Outages mailing list