Bad Hop/Routing Issue

[tj007s13]

Weeeeeeeeeeeeeeeeeeee

[Senad]

tj007s13,


Just by looking at the hop 14 and then 15 that the hop is normal when you reach the server.

The reason that this hop is so high is because the core switches prioritize tcp and udp traffic over icmp. Therefore any ICMP traffic that comes to it will be lower in preference and will have such a high ping/latency to that specific hop.

If you look after that hop the traffic ping is normal. If you look at hop 6 and hop 8 the ISPs are pretty much doing the exact same thing (IE: Prioritizing the traffic so ICMP is responded to later).

[tj007s13]

Weeeeeeeeeeeeeeeeeeeeeeeeeeeee

[Karl]

That 13 and 14 are both bad their...

Correct, that is not good. When did that happen? It seems there was a temporary issue with Savvis, right here in Chicago. Running traces now, etc. everything seems fine.

As a note, we have no control over the incoming routes, for the most part. Whichever way your ISP decides to route it is the way it is going to go. We have things in place to route around outgoing congestion, but that doesn't affect incoming routes.