|reply to Mandr4ke |
Re: Chicago Routing is WACKED! This needs to be fixed ASAP!
I'm guessing this was a temporary issue with the peering between Qwest & Savvis in Chicago, causing traffic to be routed to NY. I'm in Chicago and got this traceroute:
3 16ms 15ms 15ms 220.ge-0-1-0.cr2.chi1.speakeasy.net [188.8.131.52]
4 14ms 14ms 15ms chx-edge-01.inet.qwest.net [184.108.40.206]
5 14ms 13ms 14ms cer-core-02.inet.qwest.net [220.127.116.11]
6 14ms 13ms 13ms chp-brdr-02.inet.qwest.net [18.104.22.168]
7 14ms 13ms 12ms ber1-ge-7-3.chicagoequinix.savvis.net [22.214.171.124]
8 12ms 12ms 12ms bpr1-ge-7-0-0.chicagoequinix.savvis.net [126.96.36.199]
9 14ms 16ms 13ms bhr1-pos-0-0.ElkGrovech3.savvis.net [188.8.131.52]
The routing through Quest on the 3rd and 4th Hop:
4 chx-edge-01.inet.qwest.net (184.108.40.206) 54.999 ms 11.455 ms 10.802 ms
5 cer-core-02.inet.qwest.net (220.127.116.11) 11.305 ms 11.447 ms 10.835 ms
has not changed muched, i'm still getting 30-40 ms connecting to Local Chicago Servers from chicago, where as for the past 5 years i've been getting 8-12 ms ping time. this is really got me upset on this awful routing. If quest using interleaving or error checking or something, thats the only things that jumps ping times up like this.
There was a problem with the Cogent network yesterday that seemed to trickle down to Cogent's other peers. One of the peers that had problems was Savvis.
After noticing weird Cogent routing problems yesterday I checked with internet health report. I have not seen that much red since the level3 fiasco.
I still don't know what happened out there with Cogent. On the 14th of February, the fiber had become severed in New York.
Des Plaines, IL
The word on the street here in the Chicago colo market is that Cogent might have the best prices, but at the expense of reliability. Call this an extreme example.
I'd like to know what freak of networking caused what should have been a 16 block hop from CHX to Equinix to suddenly head for New York, going right past Equinix along the way. That area is crazy-full of fiber. That is truly screwy!
|reply to Mandr4ke |
That increase on hop 4 doesn't indicate it problem, it just indicates that Qwest's router is likely deprioritizing ICMP ping packet requests (which is the way many routers are configured). Does the full trace show higher than normal latency at the destination?