For the last couple weeks there has been a route stuck in the Chicago wan/core that is directing some Minnesota-bound traffic through Denver, even though Chicago and the Roseville, MN aggregation remain up and directly connected. This has the dual benefit of unnecessarily increasing the load on Comcast's internal backbone as well as increasing latency for Minnesota subscribers connecting to "east of the Mississippi" destinations by ~20ms.
I'm hoping Comcast engineers read this list, or someone in the carrier community can help poke one of their Comcast contacts to help get this resolved.
Thanks in advance!
"Wedged" route - 76.113.128.0/17
Correct route - 69.180.128.0/18
Example trace from Chicago source to 76.113.128.0/17:
Thanks for the folks who looked at this -- things are looking better this morning:
traceroute to 76.113.128.1 (76.113.128.1), 30 hops max, 40 byte packets
1 69.65.40.62 (69.65.40.62) 0.858 ms 0.840 ms 0.838 ms
2 so2-0-0-0.er1.Chi1.Servernap.net (69.39.239.169) 1.876 ms 1.878 ms 1.875 ms
3 ge-6-20.car1.Chicago1.Level3.net (4.79.65.49) 1.854 ms 1.858 ms 1.855 ms
4 ae-2-54.edge1.Chicago2.Level3.net (4.68.101.115) 60.047 ms 60.068 ms 60.067 ms
5 COMCAST-IP.edge1.Chicago2.Level3.net (4.71.248.26) 3.045 ms 3.051 ms 3.049 ms
6 te-0-2-0-5-ar03.roseville.mn.minn.comcast.net (68.87.174.73) 12.172 ms 12.267 ms 12.250 ms
7 te-2-1-ur01.sims.mn.minn.comcast.net (68.87.174.74) 11.717 ms * *
8 te-8-3-ur02.sims.mn.minn.comcast.net (68.87.174.78) 11.940 ms * *
9 te-2-1-ur01.newport.mn.minn.comcast.net (68.87.174.82) 12.224 ms * *
10 c-76-113-128-1.hsd1.mn.comcast.net (76.113.128.1) 12.203 ms 12.203 ms 12.045 ms
-Eric
Eric Spaeth wrote: