About once or twice a day, my tunnel twinbee-1.tunnel.tserv4.nyc4.ipv6.he.net loses connectivity for a minute or so. During those times I am able to ping6 the server endpoint, but a traceroute reveals that the packets don't get any further than that.
I assume that if this were an issue with my connectivity or my configuration, I would be unable to reach the other end of the tunnel. (Is that a reasonable assumption?) I'd prefer to verify that the issue isn't on my end before I officially bug HE support about it.
Has anyone with tunnels on the same server experienced the same issue? Thanks.
Quote from: twinbee on October 02, 2012, 02:03:48 PMI assume that if this were an issue with my connectivity or my configuration, I would be unable to reach the other end of the tunnel. (Is that a reasonable assumption?)
From your description it does sound like the tunnel server has a connectivity problem with the rest of the HE network.
Quote from: twinbee on October 02, 2012, 02:03:48 PMHas anyone with tunnels on the same server experienced the same issue?
I don't have any tunnels on that tunnel server. I'll try a periodic traceroute towards your IP to see if I can capture an output of what the situation looks like from the other end when the problem occurs.
This is what the traceroute from my end looks like right now, where the connection is working:
traceroute to twinbee-1-pt.tunnel.tserv4.nyc4.ipv6.he.net (2001:470:1f06:13b::2), 30 hops max, 80 byte packets
1 2001:470:28:940:5d75:c1f4:e0a0:f8ec 0.668 ms 1.098 ms 2.401 ms
2 2001:470:27:940::1 35.284 ms 39.439 ms 45.505 ms
3 2001:470:0:11e::1 49.641 ms 40.723 ms 41.607 ms
4 2001:470:0:22f::1 56.143 ms 60.358 ms 65.449 ms
5 2001:470:0:3f::1 72.558 ms 69.322 ms 70.277 ms
6 2001:470:0:128::1 139.692 ms 142.498 ms 149.447 ms
7 2001:470:20::2 155.211 ms 141.553 ms 145.841 ms
8 2001:470:1f06:13b::2 153.180 ms 166.884 ms 155.562 ms
Thanks, much obliged!
I never see any problems, but it's my home tunnel and I'm not constantly on it. How long does the outage last?
Around 1:12 CEST I saw an outage lasting 1-2 minutes:traceroute to twinbee-1-pt.tunnel.tserv4.nyc4.ipv6.he.net (2001:470:1f06:13b::2), 30 hops max, 80 byte packets
1 2001:470:28:940:5d75:c1f4:e0a0:f8ec 0.560 ms 26.473 ms 26.992 ms
2 2001:470:27:940::1 130.609 ms 111.741 ms 111.732 ms
3 2001:470:0:11e::1 117.794 ms 27.059 ms 30.887 ms
4 2001:470:0:22f::1 75.144 ms 76.812 ms 76.789 ms
5 2001:470:0:3f::1 71.083 ms 71.530 ms 71.941 ms
6 2001:470:0:128::1 127.003 ms 132.490 ms 129.562 ms
7 * * *
8 * * *
9 * * *
10 * * *
11 * * *
12 * * *
13 * * *
14 * * *
15 * * *
16 * * *
17 * * *
18 * * *
19 * * *
20 * * *
21 * * *
22 * * *
23 * * *
24 * * *
25 * * *
26 * * *
27 * * *
28 * * *
29 * * *
30 * * *
That was the only outage lasting more than 30 seconds. At 6:22 I saw this:traceroute to twinbee1-pt.tunnel.tserv4.nyc4.ipv6.he.net (2001:470:1f06:13b::2), 30 hops max, 80 byte packets
1 2001:470:28:940:5d75:c1f4:e0a0:f8ec 0.640 ms 26.630 ms 29.656 ms
2 2001:470:27:940::1 88.500 ms 68.132 ms 68.930 ms
3 2001:470:0:11e::1 71.353 ms 31.957 ms 35.318 ms
4 2001:470:0:22f::1 59.966 ms 65.083 ms 64.553 ms
5 2001:470:0:3f::1 81.507 ms 78.830 ms 79.176 ms
6 2001:470:0:128::1 125.480 ms 137.253 ms 134.440 ms
7 * 2001:470:20::2 3013.035 ms *
8 * * *
9 * * *
10 * * *
11 * * *
12 * * *
13 * * *
14 * * *
15 * * *
16 * * *
17 * * *
18 * * *
19 * * *
20 * * *
21 * * *
22 * * *
23 * * *
24 * 2001:470:1f06:13b::2 2979.384 ms 2859.134 ms
After that I saw multiple short outages, which appeared to be on the IPv4 side. The first one was at 7:19traceroute to twinbee-1-pt.tunnel.tserv4.nyc4.ipv6.he.net (2001:470:1f06:13b::2), 30 hops max, 80 byte packets
1 2001:470:28:940:5d75:c1f4:e0a0:f8ec 0.549 ms 21.125 ms 21.693 ms
2 2001:470:27:940::1 36.378 ms 24.820 ms 29.337 ms
3 2001:470:0:11e::1 32.421 ms 26.103 ms 39.914 ms
4 2001:470:0:22f::1 57.258 ms 58.040 ms 53.628 ms
5 2001:470:0:3f::1 63.488 ms 64.182 ms 68.110 ms
6 2001:470:0:128::1 132.173 ms 145.844 ms 132.615 ms
7 2001:470:20::2 133.246 ms 124.760 ms 128.237 ms
8 * * *
9 * * *
10 * * *
11 * * *
12 * * *
13 * * *
14 * * *
15 * * *
16 * * *
17 * * *
18 * * *
19 * * *
20 * * *
21 * * *
22 * * *
23 * * *
24 * * *
25 * * *
26 * * *
27 * * *
28 * * *
29 * * *
30 * * *
There clearly were two outages on the IPv6 side. And combined with your observations, I think it would be appropriate to email ipv6@he.net about it. There also appears to have been a few outages on the IPv4 side. With the information I have access to I cannot say if those outages are on the HE side, on your side, or somewhere in between. It is not impossible that there is one root cause responsible for both sorts of outages.
I just noticed that I still have the command running to periodically ping your tunnel. The brief outages still appear to be happening. If it is any use to you, I can attach the full output to this thread or email it to you, it is 90MB (compressed to 5MB with gzip).
Metrics involving the tservs at either nyc4 or tor1 right now are going to be unusual, as tor1 is acting as the nyc4 tserv right now, and taking more tunnels than it ever normally should.
nyc4 (and by extension tor1) should be back to normal in another week or so, time specifics variable due to Sandy-effects.
Quote from: kcochran on November 02, 2012, 10:12:56 AMtor1 is acting as the nyc4 tserv right now
The issue had been ongoing before that. But it may not make sense to investigate that issue before nyc4 is back. I'll just leave the command running for now.
tserv4.nyc4 is back in normal operation on new hardware.
Quote from: kcochran on November 16, 2012, 11:21:00 AM
tserv4.nyc4 is back in normal operation on new hardware.
Guys,
I'd been wondering about the latency increase to this tunnel... and it's now back to where it was. Thanks very much! Life is good again...
_KMP