Roadrunner's packet loss and very high latency

[bill]

This only happens to some sites I use on RR. To get a background, before July of 1999, we on Roadrunner of Maine were directly on the sprintlink backbone. Access was very fast all the time, 70ms or less to the west coast, downloads always over 100KB/sec. Then some dumbass management decided we had to get the overloaded RR public backbone for most of our Internet traffic. The sprintlink backbone is still online, as is a new AT&T one that went in in October, however the majority of our traffic is now routed over the RR backbone. So this results in problems because of the RR backbone's piss poor peering with other providers. Take a look at this..

Tracing route to audioreview.com [216.35.140.199]
over a maximum of 30 hops:

1 7 ms * 8 ms tas14-qfe0.maine.rr.com [24.31.132.1]
2 8 ms 7 ms 7 ms ubr07-atm-tas14.maine.rr.com [24.25.160.81]
3 9 ms 8 ms 8 ms ubr07-gsr01.maine.rr.com [24.25.160.49]
4 9 ms 9 ms 8 ms gsr01-pwmcore01.maine.rr.com [24.25.160.34]
5 9 ms 20 ms 9 ms pwmcore02-100bt.maine.rr.com
[204.210.65.253]
6 19 ms 10 ms 10 ms prbrt01-prrtr01.rr.com [24.218.188.30]
7 12 ms 12 ms 12 ms bsgsr01-prbrt01.rr.com [24.218.188.26]
8 20 ms 19 ms 18 ms nygsr01-bsgsr01.rr.com [24.218.188.22]
9 24 ms 29 ms 27 ms pnbrt01-nybrt01.rr.com [24.218.188.9]
10 34 ms 35 ms 36 ms pnnj-01.core.exodus.net [192.157.69.25]
11 24 ms 23 ms 24 ms bbr02-s2-3.jrcy01.exodus.net [209.1.169.209]
12 * 2702 ms * bbr02-p5-0.okbr01.exodus.net
[216.32.132.109]
13 2763 ms * * bbr02-g6-0.okbr01.exodus.net
[216.32.173.234]
14 2292 ms 2312 ms 2261 ms bbr02-p0-0-0.sntc04.exodus.net
[216.32.132.150]
15 * 2553 ms 2614 ms dcr01-g2-0.sntc04.exodus.net [216.34.2.33]
16 2376 ms 2710 ms 2743 ms rsm03-vlan920.sntc04.exodus.net
[216.34.2.67]
17 2598 ms 2625 ms * audioreview.com [216.35.140.199]
18 * 2264 ms 2788 ms audioreview.com [216.35.140.199]

Trace complete.

It looks like exodus's fault right? Nope. I emailed exodus, who did a reverse traceroute back to me, and the route back to me is different, goes over a mediaone router that's severely overloaded. So now it is RR's fault. I finally got them to admit it last friday(this has been an ongoing saga since September), and they said "there is a ticket open on this issue". Haven't heard anything since then. I doubt this problem is going to be fixed anytime soon. Quite a few sites I access are affected by this problem, and with over 30% packet loss and latency in excess of 2000ms during the 2pm to late night timeframe on weekdays and 24 hours a day on weekends, the sites are effectively unreachable. I wish Bellatlantic and Maine telephone would deploy DSL already.