mcalhoun at iodatacenters.com
Wed Feb 18 18:25:02 UTC 2009
While I agree with all of your assessments, this traceroute was being provided to illustrate where traffic *appears* to be stopping when we are seeing the issue. It's intermittent, so some times we can reach the destination hosts (via HTTP, HTTPS, etc.) and sometimes we can't.
When we can reach the destination hosts (via HTTP), the traceroute completes
When we can't reach the destination hosts (via HTTP), the traceroute won't complete and the last hop is the host that I indicated in my previous post (22.214.171.124)
From: Justin Krejci [mailto:jkrejci at usinternet.com]
Sent: Wednesday, February 18, 2009 11:15 AM
To: kgasso at visp.net; Calhoun, Matthew
Cc: 'NANOG list'
Subject: RE: McAfee/AT&T Issue
We've also seen that busy routers are slower to respond to requests directed
at them as opposed to traffic routing thru them which can continue to work
without issue or performance loss.
From: Kameron Gasso [mailto:kgasso-lists at visp.net]
Sent: Wednesday, February 18, 2009 12:03 PM
To: Calhoun, Matthew
Cc: NANOG list
Subject: Re: McAfee/AT&T Issue
Calhoun, Matthew wrote:
> 9 212 ms 200 ms * 126.96.36.199 <--------Problem occurring
here. Sometimes traffic gets through, sometimes it doesn't
> 10 29 ms 26 ms 26 ms 188.8.131.52
> 11 26 ms 26 ms 26 ms www.mcafeeasap.com [184.108.40.206]
Looks a lot like that hop is rate-limiting ICMP to itself. Everything
beyond it seems to be good from the looks of it.
More information about the NANOG