[Outages-discussion] [outages] GTT daily congestion in Europe

[Brian Turnbow]

Hi,

> > The congestion might be on the reverse path.
> > Can you get a trace route in the opposite direction?
> 
> Moving to -discussions.
> 
> I can exclude that because:
> 
> - I did fully analyze the issue, including clarifying the direction with reverse
> MTRs
> - some traffic was rerouted on the opposite direction: ZERO success
> - some traffic was rerouted on the mention direction: FULL success
> 
> I did not share every single data point available to me in this post, because that
> was not the point.

I can confirm.
Example from Mil to Fra
If sent using loopback address  with return via NTT all is good

RP/0/RP0/CPU0:NCS-55A2-CALDERA-1#trace 85.10.242.181 source l0  
Fri Feb  4 16:01:21.446 CET

Type escape sequence to abort.
Tracing the route to 85.10.242.181

 1  xe-0-2-3.cr2-mil3.ip4.gtt.net (46.33.66.5) 1 msec  1 msec  1 msec
 2  ae9.cr6-fra2.ip4.gtt.net (141.136.110.41) 16 msec  16 msec  16 msec
 3  ip4.gtt.net (154.14.78.154) 20 msec  19 msec  20 msec
 4  core11.nbg1.hetzner.com (213.239.245.249) 22 msec  23 msec  22 msec
 5  spine12.cloud1.nbg1.hetzner.com (213.239.239.142) 23 msec
    spine12.cloud1.nbg1.hetzner.com (213.239.239.122) 24 msec
    spine12.cloud1.nbg1.hetzner.com (213.239.239.142) 23 msec
 

using the interface IP that is GTT space with return via GTT we can easily see delay.

RP/0/RP0/CPU0:NCS-55A2-CALDERA-1#trace 85.10.242.181
Fri Feb  4 16:01:36.548 CET

Type escape sequence to abort.
Tracing the route to 85.10.242.181

 1  xe-0-2-3.cr2-mil3.ip4.gtt.net (46.33.66.5) 1 msec  2 msec  1 msec
 2  ae9.cr6-fra2.ip4.gtt.net (141.136.110.41) 22 msec  17 msec  16 msec
 3  ip4.gtt.net (154.14.78.154) 45 msec  57 msec  63 msec
 4  core12.nbg1.hetzner.com (213.239.245.253) 66 msec  67 msec  69 msec
 5  spine12.cloud1.nbg1.hetzner.com (213.239.239.122) 101 msec
    spine12.cloud1.nbg1.hetzner.com (213.239.239.142) 71 msec

Brian