Tickets - Bad ping times with two different dedus01 tunnels :: SixXS

Ticket ID: SIXXS #996277
Ticket Status: User

PoP: dedus01 - SpeedPartner GmbH (Duesseldorf)

Bad ping times with two different dedus01 tunnels

Shadow Hawkins on Wednesday, 11 March 2009 01:05:48

I have read and followed the "Reporting Problems" section on the Contact page and am providing the following details for this report based on the list of items stated there: Hello all, first some information about me and my tunnels: User: MSV6-SIXXS Heartbeat tunnel: T16455 Static tunnel: T18281 PoP: dedus01 OS used at tunnel endpoints: Debian with kernel 2.6.x Pings to the "outside" IPv6 world (I tried several IPv6 hosts in different countries and different providers) show that the connection is not very constant or permanent. I got good pings 5-15 times and then I got a response time from between 200 and 1000ms. It's noticable also when using PuTTY to the remote machine for example. I attached two ping logs below. Loss statistics here are just fine, nothing to complain about. Pings inside dedus01 world are great and show nothing special. I got this issue for over three months with two dedus01 tunnels, one is a heartbeat tunnel, the other is a static tunnel. Everything else is up and running fine as you can see in my tunnel statistics. Here's an example:

64 bytes from atlan.shadowland-network.de: icmp_seq=1 ttl=53 time=70.8 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=2 ttl=53 time=263 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=3 ttl=53 time=26.7 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=4 ttl=53 time=28.1 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=5 ttl=53 time=27.2 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=6 ttl=53 time=26.3 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=7 ttl=53 time=27.6 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=8 ttl=53 time=26.4 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=9 ttl=53 time=26.5 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=10 ttl=53 time=27.5 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=11 ttl=53 time=27.4 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=12 ttl=53 time=24.2 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=13 ttl=53 time=26.5 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=14 ttl=53 time=26.3 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=15 ttl=53 time=28.0 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=16 ttl=53 time=25.9 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=17 ttl=53 time=478 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=18 ttl=53 time=27.0 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=19 ttl=53 time=26.4 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=20 ttl=53 time=26.9 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=21 ttl=53 time=25.5 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=22 ttl=53 time=25.7 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=23 ttl=53 time=26.8 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=24 ttl=53 time=26.6 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=25 ttl=53 time=25.4 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=26 ttl=53 time=27.3 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=27 ttl=53 time=438 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=28 ttl=53 time=26.5 ms 64 bytes from atlan.shadowland-network.de: icmp_seq=29 ttl=53 time=30.4 ms

Is this known? Can anybody else who also uses dedus01 tunnels confirm this? It's very simple to test from any machine behind the tunnel endpoint. Regards, Martin

State change: user Locked

Jeroen Massar SixXS Staff

on Wednesday, 11 March 2009 01:13:52

The state of this ticket has been changed to user

Bad ping times with two different dedus01 tunnels

Jeroen Massar SixXS Staff

on Wednesday, 11 March 2009 01:14:25

Please try to look also at the PATH of this connection, both IPv4 (as you are tunneling) and IPv6.

Bad ping times with two different dedus01 tunnels

Shadow Hawkins on Wednesday, 11 March 2009 10:18:20

With path you mean routing tables? If yes, here's the one from my Linux software router on the heartbeat tunnel:

Kernel IP routing table Destination Gateway Genmask Flags Metric Ref Use Iface 172.27.0.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0 85.216.56.0 0.0.0.0 255.255.252.0 U 0 0 0 eth1 0.0.0.0 85.216.56.1 0.0.0.0 UG 0 0 0 eth1

Kernel IPv6 routing table Destination Next Hop Flag Met Ref Use If 2a01:198:200:18e::/64 :: Un 256 0 84 sixxs 2a01:198:2b4::/64 :: U 256 0 0 eth0 2000::/3 2a01:198:200:18e::1 UG 1024 0 5292 sixxs fe80::/64 :: U 256 0 0 eth0 fe80::/64 :: U 256 0 0 eth1 fe80::/64 :: Un 256 0 0 sixxs ::/0 2a01:198:200:18e::1 UG 1024 0 0 sixxs ::/0 :: !n -1 1 13111 lo ::1/128 :: Un 0 1 26 lo 2a01:198:200:18e::/128 :: Un 0 1 0 lo 2a01:198:200:18e::2/128 :: Un 0 1 153 lo 2a01:198:2b4::/128 :: Un 0 1 0 lo 2a01:198:2b4::1/128 :: Un 0 1 41 lo fe80::/128 :: Un 0 1 0 lo fe80::/128 :: Un 0 1 0 lo fe80::/128 :: Un 0 1 0 lo fe80::55d8:3b85/128 :: Un 0 1 0 lo fe80::ac1b:1/128 :: Un 0 1 0 lo fe80::202:ffff:fe00:b55a/128 :: Un 0 1 0 lo fe80::2d0:b7ff:fe9f:c59b/128 :: Un 0 1 227 lo ff00::/8 :: U 256 0 0 eth0 ff00::/8 :: U 256 0 0 eth1 ff00::/8 :: U 256 0 0 sixxs ::/0 :: !n -1 1 13111 lo

And here the machine with the static tunnel:

Kernel IP routing table Destination Gateway Genmask Flags Metric Ref Use Iface 0.0.0.0 81.169.128.1 0.0.0.0 UG 0 0 0 eth0

Kernel IPv6 routing table Destination Next Hop Flag Met Ref Use If 2a01:198:200:29b::/64 :: Un 256 0 13968 v6sixxs 2a01:198:32b::/64 :: U 256 0 0 eth0 2a01:198:32b::/64 :: Un 256 0 0 v6sixxs 2000::/3 2a01:198:200:29b::1 UG 1024 0227694 v6sixxs fe80::/64 :: U 256 0 0 eth0 fe80::/64 :: Un 256 0 0 v6sixxs ::/0 :: !n -1 1241663 lo ::1/128 :: Un 0 1 6456 lo 2a01:198:200:29b::2/128 :: Un 0 1 65902 lo 2a01:198:32b::1/128 :: Un 0 13007183 lo 2a01:198:32b::1/128 :: Un 0 1 61025 lo fe80::51a9:835e/128 :: Un 0 1 0 lo fe80::221:85ff:fefa:8f86/128 :: Un 0 1 10 lo ff00::/8 :: U 256 0 0 eth0 ff00::/8 :: U 256 0 0 v6sixxs ::/0 :: !n -1 1241663 lo

Regards, Martin

Bad ping times with two different dedus01 tunnels

Jeroen Massar SixXS Staff

on Wednesday, 11 March 2009 21:39:13

With path you mean routing tables

Traceroute, or maybe better 'mtr' will show you where the packets get slowed-down. Then again, if your IPv4 ISP does ratelimitting^WQuality of Service on 'unknown' or 'odd' packets, then pings might not show any latency increase, while it will show up on the actual protocol.

Bad ping times with two different dedus01 tunnels

Shadow Hawkins on Wednesday, 11 March 2009 22:41:13

Ok here you are: heartbeat tunnel:

Packets Pings Host Loss% Snt Last Avg Best Wrst StDev 1. gw-399.dus-01.de.sixxs.net 0.2% 478 16.8 17.3 13.3 44.0 3.1 2. 2a01:198:4:6::1 0.6% 477 15.1 33.4 13.6 515.7 59.7 3. speedpartner.r1.dus1.de.opencarrier.eu 1.9% 477 16.1 32.7 13.5 547.6 58.5 4. 2001:7f8:3a:e003::2 0.6% 477 22.3 36.8 17.5 489.9 49.0 5. de-cix20.net.google.com 1.5% 477 20.7 41.0 19.0 491.4 54.9 6. 2001:4860::23 1.3% 477 25.6 36.2 19.8 334.9 42.8 7. 2001:4860:a003::68 1.0% 477 190.7 37.7 19.2 501.4 58.3

static tunnel:

Packets Pings Host Loss% Snt Last Avg Best Wrst StDev 1. gw-668.dus-01.de.sixxs.net 0.0% 542 13.9 14.9 13.6 41.2 2.3 2. 2a01:198:4:6::1 0.7% 542 15.5 24.3 14.1 449.3 42.0 3. speedpartner.r1.dus1.de.opencarrier.eu 1.1% 542 275.8 28.8 14.2 548.2 59.3 4. oc-r1-dus1.r1.fra3.de.opencarrier.eu 0.7% 542 183.8 35.2 18.1 537.4 57.9 5. te-2-4.bb-d.fra3.fra.de.oneandone.net 0.7% 542 91.8 36.3 19.1 551.1 56.9 6. te-1-4.bb-c.act.fra.de.oneandone.net 0.9% 542 20.7 33.0 19.0 540.0 53.9 7. te-3-3.bb-c.bs.kae.de.oneandone.net 0.7% 542 21.7 35.6 21.1 526.8 54.5 8. ge-0-0-0-0.rr-a.bap.rhr.de.oneandone.net 1.7% 542 23.6 36.5 22.2 441.4 50.0 9. te-1-2.gw-dista-a.bs.ka.oneandone.net 1.1% 541 23.7 40.0 22.1 553.4 57.9 10. 2001:8d8:0:6::12 1.3% 541 24.1 38.9 23.0 479.5 53.7 11. proteus.hot-chilli.net 1.1% 541 24.1 37.4 23.4 387.6 46.2

As you can see the higher latencies start at dedus01, not on the way to the tunnel provider. As I run the tunnels from different ISPs (heartbeat with KabelBW which is cable internet and the static tunnel from a LAN connection) I do not think that ratelimiting on IPv4 is the problem. Regards, Martin

Bad ping times with two different dedus01 tunnels

Shadow Hawkins on Saturday, 14 March 2009 01:16:48

Hm, any idea? For me it looks like a provider problem, as the pings from my machines to dedus01 are just fine.

Bad ping times with two different dedus01 tunnels

Shadow Hawkins on Wednesday, 11 March 2009 10:22:52

Some more information: heartbeat tunnel: Tunnel endpoint IP: 2a01:198:200:18e::2 Prefix: 2a01:198:2b4::/48 static tunnel: Tunnel endpoint IPv6: 2a01:198:200:29b::2 IPv4: 81.169.131.94 Prefix: 2a01:198:32b::/48

Bad ping times with two different dedus01 tunnels

Shadow Hawkins on Monday, 16 March 2009 22:55:20

Hello? Can I do anything else? I want to solve this but cannot to this alone, especially because I think the problem is not with my hard- or software. Regards, Martin

Bad ping times with two different dedus01 tunnels

Shadow Hawkins on Thursday, 16 July 2009 12:24:18

Works now... Did not change anything on the affected machines. :-)

Bad ping times with two different dedus01 tunnels

Jeroen Massar SixXS Staff

on Thursday, 16 July 2009 13:08:16

Ever thought about the fact that things might not happen the direct instant that you change something?

Please note Posting is only allowed when you are logged in.