[outages] flapping issues reaching East Coast AWS

[William Kern]

We have seen flapping all morning reaching our AWS East Coast servers 
from our Thousand Oaks location. They drop off for a minute or so and 
then come back.

It seems to occur at 3-5 minute intervals for several cycles, then 
everything is fine for 30 minutes or so. Then the alarms start to come 
in again so wash rinse and repeat.

The issue is definately not local, as our local monitoring to our peer 
connections are fine as are our connections to AWS WestCoast servers. 
The only complaint I got so far was from a customer on the East cost 
trying to VPN to a server here and getting booted out a lot.

Traceroutes imply there may be an issue with Frontier as the few times I 
have caught the problem while logged into the remote East Coast server 
the path outbound from AWS East goes to Dallas and then seems to have 
issue reaching Los Angeles. I also show those traceroute pages moving 
around a bit between those locations.

The connection from here to the the East Cost show weirdness involving 
Charter so maybe the issue is on the reverse trip.

this is pretty typical.

  8  lag-401.dllstx976iw-bcr00.netops.charter.com (66.109.5.229) 36.002 
ms  36.316 ms lag-16.dllstx976iw-bcr00.netops.charter.com (66.109.6.1)  
36.451 ms
  9  lag-0.pr3.dfw10.netops.charter.com (66.109.5.121)  39.187 ms 
lag-302.pr3.dfw10.netops.charter.com (209.18.43.77)  42.279 ms 45.338 ms
10  99.83.71.242 (99.83.71.242)  35.427 ms 99.83.71.240 (99.83.71.240)  
36.389 ms 99.82.176.170 (99.82.176.170)  35.339 ms
11  * 150.222.206.169 (150.222.206.169)  34.071 ms *
12  * * 15.230.48.42 (15.230.48.42)  33.717 ms

Finally its not within AWS as our WestCoast servers can maintain a 
steady ping with the affected East Coast sites.

So is anyone else seeing weirdness on cross country trips?


William Kern

PixelGate