London Tunnel increase in latency?

[mhisani]

Hello first post. Apologies if I post this in the wrong place however. I setup a tunnel via yourselves and set it up on my DDWRT router. All went well (This is my "get used to ipv6 toy) Now up until a few days ago the tunnel had great latency however all of a sudden it has changed as you can se here..



From what I can tell its bidirectional. Latency has gone from around 20 to 100-130.

The tunnel is setup on London as I live pretty close. I apologise for not being a true genius about ipv6 but i'm learning.

If anyone could give me a little help I'd be grateful.

[kasperd]

There are two commands, that can help you narrow down the reason for the latency increase. Try to run a traceroute from your network towards the IPv4 address of the tunnel server. Additionally try to run a traceroute from your network towards some IPv6 address outside of the HE network.

[mhisani]

Tracing route to tserv1.lon1.he.net [216.66.80.26]
over a maximum of 30 hops:

  1    <1 ms    <1 ms    <1 ms  feral [10.0.0.1]
  2    18 ms     7 ms     8 ms  10.153.48.1
  3    10 ms    13 ms     7 ms  cmbg-core-1b-ae2-3971.network.virginmedia.net [6
2.253.128.225]
  4    20 ms    28 ms    42 ms  nrth-bb-1b-ae5-0.network.virginmedia.net [212.43
.163.145]
  5    13 ms    19 ms    11 ms  tele-ic-4-ae0-0.network.virginmedia.net [62.253.
174.18]
  6    92 ms   124 ms    99 ms  10gigabitethernet2-2.core1.ash1.he.net [206.223.
115.37]
  7   101 ms   104 ms   138 ms  10gigabitethernet1-2.core1.nyc4.he.net [72.52.92
.85]
  8    87 ms   115 ms   106 ms  10gigabitethernet1-2.core1.lon1.he.net [72.52.92
.242]
  9   138 ms    88 ms    87 ms  tserv1.lon1.he.net [216.66.80.26]
Trace complete.


----

Tracing route to wyrddreams.org [2001:41c8:1:6081::92]
over a maximum of 30 hops:

  1    <1 ms    <1 ms    <1 ms  2001:470:1f09:1d83::11
  2     *        *        *     Request timed out.
  3   100 ms   103 ms    99 ms  gige-g4-8.core1.lon1.he.net [2001:470:0:67::1]
  4    92 ms    95 ms    90 ms  2001:7f8:4::8a61:1
  5   120 ms   126 ms   101 ms  2001:41c8:0:76::5
  6   111 ms   101 ms   113 ms  2001:41c8:0:100::2
  7    97 ms    99 ms    99 ms  2001:41c8:0:95::1
  8    99 ms   100 ms    98 ms  jamie.bytemark.co.uk [2001:41c8:0:858::35]
  9   112 ms   109 ms   101 ms  2001:41c8:1:6081::92


Is this showing its going from the UK to NYC then back to the UK?

[kasperd]

Is this showing its going from the UK to NYC then back to the UK?

I suppose it does, though I don't know the exact physical location of each of the routers on the path.

Looks like most of your latency is on the hop between tele-ic-4-ae0-0.network.virginmedia.net [62.253.174.18] and 10gigabitethernet2-2.core1.ash1.he.net [206.223.115.37] (where in the world is ash1?)

Having the latency show up on a peering link makes it a bit harder for you to track down who is responsible for the latency. Maybe the peering link your packets used to go through is no more. Maybe the advertisements from HE changed and caused traffic to take a suboptimal path. Maybe virginmedia is routing traffic through a suboptimal path.

I can think of two ways for you to proceed. Since you appear to be a customer of virginmedia (directly or indirectly), you can contact them and ask what's up with that? You'd have go through the customer support at your ISP (which AFAIK sucks for most ISPs). Alternatively you can aim for the other end and contact ipv6@he.net and ask what may be the problem.

In any case you'd need to include the output of your IPv4 traceroute and ask why the traffic takes such a long detour.

[broquea]

Looks like Virgin UK peering bounced and didn't re-establish. They should fix that :) (They being Virigin)

ASH = Ashburn, VA

[aamkeri]

If i understand this problem correctly when the traffic is traversing through the tunnel then we are seing the issue. Can you please check are we seeing any errors on hub or spoke side like replay errors or not and is there any qos applied on the spoke or hub side .