I'm seeing some routing weirdness from PlusNet xDSL connections to one of our subnets, and wondered if there was any members from PlusNet and Virgin on list who could take a look.
We have a subnet 195.2.224.0/24 on AS44052 which is dual homed to EntaNet and IXReach. >From a PlusNet xDSL connection, the routing to our router looks quite sensible when hitting the carrier side IP addresses that are assigned to it: EntaNet assigned address: alan-mbp:~ alanramsay$ traceroute -I 188.39.202.154 traceroute to 188.39.202.154 (188.39.202.154), 64 hops max, 72 byte packets 1 nat (172.16.10.1) 5.767 ms 3.174 ms 1.931 ms 2 410.xe-2-2-0.central10.pcn-bng02.plus.net (84.93.253.72) 17.874 ms 17.056 ms 14.640 ms 3 411.be6.pcn-ir02.plus.net (84.93.253.79) 17.152 ms 14.379 ms 17.687 ms 4 195.99.125.144 (195.99.125.144) 14.960 ms 17.165 ms 14.512 ms 5 peer1-et-10-3-0.telehouse.ukcore.bt.net (62.172.103.17) 18.485 ms 16.459 ms 14.942 ms 6 * * * 7 te0-2-0-3.interxion3.core.enta.net (188.39.127.172) 18.420 ms 18.229 ms 17.572 ms 8 te0-0-2-0.miltonkeynes.core.enta.net (188.39.127.209) 19.556 ms 18.382 ms 20.486 ms 9 te0-0-0-1.birmingham-te.core.enta.net (188.39.127.212) 22.215 ms 30.097 ms 21.483 ms 10 te0-0-0-2.wolv-te.core.enta.net (188.39.127.217) 22.330 ms 25.702 ms 21.107 ms 11 188-39-202-154.static.enta.net (188.39.202.154) 20.173 ms 21.385 ms 20.329 ms alan-mbp:~ alanramsay$ IXReach assigned address: alan-mbp:~ alanramsay$ traceroute -I 31.217.134.34 traceroute to 31.217.134.34 (31.217.134.34), 64 hops max, 72 byte packets 1 nat (172.16.10.1) 5.455 ms 2.448 ms 2.362 ms 2 410.xe-2-2-0.central10.pcn-bng02.plus.net (84.93.253.72) 16.839 ms 14.814 ms 14.290 ms 3 411.be6.pcn-ir01.plus.net (84.93.253.75) 15.082 ms 15.229 ms 15.825 ms 4 195.99.125.136 (195.99.125.136) 15.509 ms 15.857 ms 16.325 ms 5 62.172.103.49 (62.172.103.49) 41.430 ms 17.144 ms 15.461 ms 6 * * * 7 31.217.134.34 (31.217.134.34) 24.338 ms 21.765 ms 21.651 ms alan-mbp:~ alanramsay$ Now if we try to do a traceroute to the router directly within the advertised subnet, PlusNet seems to favour the EntaNet connection, but sends traffic to it via Virgin Media. alan-mbp:~ alanramsay$ traceroute -I 195.2.224.253 traceroute to 195.2.224.253 (195.2.224.253), 64 hops max, 72 byte packets 1 nat (172.16.10.1) 4.343 ms 2.126 ms 5.047 ms 2 410.xe-2-2-0.central10.pcn-bng02.plus.net (84.93.253.72) 17.572 ms 16.744 ms 13.716 ms 3 411.be6.pcn-ir02.plus.net (84.93.253.79) 19.078 ms 18.111 ms 15.559 ms 4 195.99.125.140 (195.99.125.140) 15.161 ms 15.049 ms 15.472 ms 5 peer2-et-1-1-0.redbus.ukcore.bt.net (62.172.103.45) 15.296 ms 15.722 ms 17.498 ms 6 194.74.65.22 (194.74.65.22) 22.802 ms 27.988 ms 23.007 ms 7 * * * 8 * * * 9 * * **10 tcl5-ic-5-ae0-0.network.virginmedia.net <http://tcl5-ic-5-ae0-0.network.virginmedia.net> (62.254.84.66) 28.122 ms 23.156 ms 24.506 ms *11 * * * 12 te0-2-0-3.interxion3.core.enta.net (188.39.127.172) 23.327 ms 20.859 ms 21.733 ms 13 te0-0-2-0.miltonkeynes.core.enta.net (188.39.127.209) 22.365 ms 23.298 ms 25.178 ms 14 te0-0-0-1.birmingham-te.core.enta.net (188.39.127.212) 26.434 ms 27.137 ms 29.711 ms 15 te0-0-0-2.wolv-te.core.enta.net (188.39.127.217) 31.469 ms 31.309 ms 28.197 ms 16 ip195-2-224-253.presscomputers.net (195.2.224.253) 28.177 ms 25.894 ms 27.156 ms alan-mbp:~ alanramsay$ What would cause the traffic to go via Virgin, rather than direct - we can see when they hit the external PlusNet IP of the router it routes without going via Virgin. Any advice on how to resolve this is much appreciated. Thanks Alan
