OT: Serious ISP problems....

Sat Oct 7 20:15:16 PDT 2006

Hello,

It's the weekend, so a real OT post won't bother many... and I would like to
tap the huge mind pool of this forum for some networking pzaaaz.

You may have noticed I have not been doing a lot on either this forum or the
new www.fptechforum.com this past couple of weeks. I've been having terrible
ISP problems. Basically, they turn the service completely off for long
periods of time like two hours and say it was for "maintenance". No warning,
no schedule, just every night between 2 and 4am. (I would love to know if
that is even legal for a large carrier, like this cable service, to do. )
Even so, that is not the problem, usually when they are having problems, I
get routed through a strange chain of routers from here in PA to our site in
NJ (only 120 miles away). Often, I go from NYC to WDC, down to Atlanta, out
to Dallas, Fort Worth than back through Indiana back to NYC.  All of this
translates to huge packet loss... usually 20% to 40% but a definite average
over the past two weeks of 33%. I talk to the techs at the NOC and they say
"It's not our problem..." ...  So, in an attempt to track down the problem
(or any problem related) I have done days of testing... when I can get
out... and I always come to the conclusion that whenever I am routed over
something called "us.above.net" routers my service is essentially unusable
with tremendous packet loss. all of their routers are pinging at above 300ms
(200ms to 800ms usually). The most notable problem is a single router which
is *always* the first router that my cable company "Metrocast" connects to
either in or out of NYC. It has the name "ny.ny.metconnect in it... and when
I nslookup it, the return leads me to "Metconnect", New York's' Free ISP".
(googled).  DIG doesn't tell me anything. I don't care what nonsense this
Free is all about.  I don't care... I just want to know what my recourse is.
Can I reach them (if so, how??) and can I tell/ask/demand that they fix this
problem router? And, assuming I could do that, *why* wouldn't my ISP have
the ability to do that, and why wouldn't they want to do that. I'm a
business user (paying big $$$) who is about to leave because of it, but
these level 2 techs keep saying "It's not our problem."  Is it time for me
to start calling VP's?

I'm at my wit's end. Here is a traceroute from Nexus to my site... going
through this *always* bad router... check hop# 14.  Then there are two pings
out of my site directly to this router. Notice the 33% packet loss.

My service is unusable, I can barely understand people on the FP talk
server... though I'm still there 24 hours a day pretty much... just having
trouble "hearing" everyone. They seem to be able to hear me most of the
time. all other VPN work is virtually impossible because of very slow
character-response times from the host. I type an "h" and it takes seconds
for it to appear on my screen.

Well, that's it. I'll take any advice on what to do technically, legally,
whatever. It is going to be VERY expensive for me to go back to a T1 or any
high-speed DSL (about $1200/month in my area). I think this cable company
"Metrocast" knows this... and therefore doesn't really care that my service
has been unusable for the past going on 3 weeks now.

Thanks in advance,

Following are the traceroute in to me... and the two pings out to *it*...

John Esak

traceroute to my site in NJ to the badly performing site in PA - HOP# 14!!!

 1  12.44.56.161 (12.44.56.161)  1.890 ms  1.761 ms  1.786 ms
 2  12.124.191.73 (12.124.191.73)  35.744 ms  214.013 ms  4.295 ms
 3  gbr2-a33s2.n54ny.ip.att.net (12.127.0.46)  5.432 ms  5.388 ms  5.458 ms
 4  tbr1-p013301.n54ny.ip.att.net (12.122.11.5)  5.782 ms  5.890 ms  6.089
ms
 5  ggr3-ge40.n54ny.ip.att.net (12.122.81.9)  4.523 ms  4.527 ms  4.496 ms
 6  att-gw.dc.aol.com (192.205.32.2)  4.791 ms  4.854 ms  4.784 ms
 7  0.ge-5-1-0.XL4.NYC4.ALTER.NET (152.63.3.121)  4.802 ms  4.900 ms  4.794
ms
 8  0.so-1-2-0.XL2.NYC1.ALTER.NET (152.63.29.101)  5.229 ms  5.117 ms  5.172
ms
 9  POS7-0.GW12.NYC1.ALTER.NET (152.63.29.197)  5.111 ms  5.090 ms  5.063 ms
10  band-x-gw.customer.alter.net (157.130.2.218)  5.264 ms  5.198 ms  5.232
ms
11  204.8.103.234 (204.8.103.234)  10.748 ms  11.091 ms  10.743 ms
12  216.194.46.86 (216.194.46.86)  10.879 ms  10.857 ms  10.796 ms
13  216.194.46.102 (216.194.46.102)  11.452 ms  11.740 ms  11.251 ms
14  216-194-30-46.ny.ny.metconnect.net (216.194.30.46)  357.849 ms  383.204
ms
385.098 ms
15  d-72-9-16-8.metrocast.net (72.9.16.8)  395.304 ms  356.698 ms  273.500
ms
16   *  *  *
17   *  *  *

What's weird is when I ping this router directly... I get normal results.

npi-p20# ping 216.194.30.46
PING 216.194.30.46 (216.194.30.46): 56 data bytes
64 bytes from 216-194-30-46.ny.ny.metconnect.net (216.194.30.46): icmp_seq=0
ttl
=239 time=312.638 ms
64 bytes from 216-194-30-46.ny.ny.metconnect.net (216.194.30.46): icmp_seq=1
ttl
=239 time=234.258 ms
64 bytes from 216-194-30-46.ny.ny.metconnect.net (216.194.30.46): icmp_seq=2
ttl
=239 time=311.409 ms
64 bytes from 216-194-30-46.ny.ny.metconnect.net (216.194.30.46): icmp_seq=3
ttl
=239 time=233.843 ms
64 bytes from 216-194-30-46.ny.ny.metconnect.net (216.194.30.46): icmp_seq=4
ttl
=239 time=265.156 ms

--- 216.194.30.46 ping statistics ---
5 packets transmitted, 5 packets received, 0% packet loss
round-trip min/avg/max = 233.843/271.461/312.638 ms

Yet when I ping out of my site to the same site of the traceroute (so its
going through this router (I must assume) to the site past the bad router
(i.e., the tracerouted site) I get very bad results of 33% packet loss...
even when the router has a couple double digit ping time as in the second
ping shown here.

PING 12.44.56.168 (12.44.56.168): 56 data bytes
64 bytes from 12.44.56.168 (12.44.56.168): icmp_seq=0 ttl=113 time=339.820
ms
64 bytes from 12.44.56.168 (12.44.56.168): icmp_seq=1 ttl=113 time=269.637
ms
64 bytes from 12.44.56.168 (12.44.56.168): icmp_seq=3 ttl=113 time=299.542
ms
64 bytes from 12.44.56.168 (12.44.56.168): icmp_seq=4 ttl=113 time=359.810
ms

--- 12.44.56.168 ping statistics ---
6 packets transmitted, 4 packets received, 33% packet loss
round-trip min/avg/max = 269.637/317.202/359.810 ms

PING 12.44.56.168 (12.44.56.168): 56 data bytes
64 bytes from 12.44.56.168 (12.44.56.168): icmp_seq=0 ttl=113 time=305.770
ms
64 bytes from 12.44.56.168 (12.44.56.168): icmp_seq=2 ttl=113 time=45.265 ms
64 bytes from 12.44.56.168 (12.44.56.168): icmp_seq=3 ttl=113 time=178.204
ms
64 bytes from 12.44.56.168 (12.44.56.168): icmp_seq=4 ttl=113 time=309.654
ms

--- 12.44.56.168 ping statistics ---
6 packets transmitted, 4 packets received, 33% packet loss
round-trip min/avg/max = 45.265/209.723/309.654 ms