Web Hosting Talk : Web Hosting Main Forums : Dedicated Server : Managed.com: weird routing issues at above.net

[MiSV]

Check this out folks. Something went horribly wrong with above.net routing.

Two machines, machine A (managed.com) and machine B (Israel)

ssh from B to A works.
ssh from A to B does not.

Traceroute from B to A:

 1  xxx  (xxx)  0.174 ms  0.165 ms  0.129 ms
 2  xxx (xxx)  0.345 ms  0.340 ms  0.297 ms
 3  bzq-115-188-65.cust.bezeqint.net (192.115.188.65)  0.365 ms  0.456 ms  0.361 ms
 4  bzq-25-90-1.cust.bezeqint.net (212.25.90.1)  0.507 ms  0.546 ms 0.636 ms
 5  pos1-2.cr01.frf02.pccwbtn.net (63.218.15.41)  78.695 ms  78.693 ms  78.580 ms
 6  pos5-1.cr02.sea01.pccwbtn.net (63.216.14.41)  238.765 ms  238.728 ms  238.697 ms
 7  six.assertivenetworks.net (198.32.180.98)  239.282 ms  238.962 ms *
 8  375.ge-6-1-0.cr1.sea1.us.above.net (209.133.64.98)  239.297 ms 239.149 ms  239.287 ms
 9  so-3-2-0.mpr3.sjc2.us.above.net (64.125.28.182)  254.673 ms 254.667 ms  254.727 ms
10  so-2-0-0.er10a.sjc2.us.above.net (64.125.30.90)  257.388 ms  257.224 ms  257.631 ms
11  reserved.above.net (209.133.64.124)  252.027 ms  252.022 ms  252.035 ms
12  assertive.managed.com (66.154.102.5)  254.942 ms  254.783 ms  254.875 ms
13  66.79.175.2 (66.79.175.2)  255.245 ms  255.543 ms  273.644 ms
14  xxx (65.75.18x.xxx)  249.685 ms  249.493 ms  249.364 ms

Looks good right? Now let's try the other way around.

 1  65.75.184.1 (65.75.184.1)  0.527 ms  0.489 ms  0.424 ms
 2  66.79.175.1 (66.79.175.1)  0.451 ms  0.314 ms  0.320 ms
 3  66-154-102-49.assertivenetworks.net (66.154.102.49)  0.378 ms 0.368 ms  0.360 ms
 4  728.ge-4-1-1.er10b.sjc2.us.above.net (209.133.64.123)  0.441 ms 0.509 ms  0.501 ms
 5  so-2-0-0.mpr4.sjc2.us.above.net (64.125.30.101)  0.449 ms  0.518 ms  0.469 ms
 6  so-3-3-0.cr1.dfw2.us.above.net (64.125.29.58)  45.352 ms  45.340 ms  45.155 ms
 7  so-0-0-0.cr2.dfw2.us.above.net (64.125.28.210)  45.260 ms  45.189 ms  45.205 ms
 8  so-2-0-0.cr2.dca2.us.above.net (64.125.29.10)  73.133 ms  73.092 ms  73.291 ms
 9  so-0-0-0.cr1.dca2.us.above.net (64.125.29.121)  73.442 ms  73.498 ms  73.527 ms
10  so-6-0-0.cr1.lhr3.uk.above.net (64.125.31.185)  144.864 ms  144.859 ms  144.802 ms
11  pos12-0.cr1.cdg2.fr.above.net (64.125.31.130)  151.862 ms  151.913 ms  152.044 ms
12  pos5-0.mpr1.cdg4.fr.above.net (208.184.231.213)  152.314 ms  152.270 ms  152.167 ms
13  * * *
14  213.223.42.46 (213.223.42.46)  153.173 ms  153.267 ms  153.099 ms
15  213.223.42.69 (213.223.42.69)  153.383 ms  153.538 ms  153.357 ms
16  213.223.42.10 (213.223.42.10)  153.308 ms  153.239 ms  153.396 ms
17  ppp-193.net-11.magic.fr (195.115.113.193)  154.084 ms !A * *
18  ppp-193.net-11.magic.fr (195.115.113.193)  154.454 ms !A * *
19  * ppp-193.net-11.magic.fr (195.115.113.193)  154.302 ms !A *
.... et cetera ad infinitum ...

So lemme get this right... packets from a high bandwith machine A sitting in US destined for another high bandwith machine B in Middle East get routed to... a dial-up user in France?

Anyone else observes oddities like this? Is this a routing issue at above.net (as traceroute suggests) or misconfigured peering at managed.com?

I paid attention to it just now (emailed their support already). It totally, mindbogglingly screws up my work environment.

So whom to flame, managed.com or above.net? Or am I stupid?

[MiSV]

Update: after short discussion with above.net it turns out that it's the French who are at fault. Unfortunately, they can't tell for sure who is responsible for maintaining the router at hop 13.

God damnit.

[Asynja]

<Insert swipe at the French here>

[Jhorra]

<Additional mumbling about the French>