[TriLUG] really slow web. help!

[Tim Jowers]

I recall Jason saying the web problems are usually DNS. Can someone tell me
what to check?

Can someone give me a tracert from a T1 here in town? I'm in Cary and can't
take this crappy network anymore.

I am on Time-Warner cable in Cary. My page load times today are about 10
seconds for big sites like Google and 30 seconds for lesser sites. It's
killing me.  Today is a little worse than normal but its always bad.

I tried BellSouth for 4 months (had both side-by-side) but the page loads
times were no better. So much for the claims DSL is better than cable. Also,
downloads from my computer to computers at relatives houses in SC were no
better. The extra I paid for faster uploads with BLS seems non-factual. The
funny part is the tests like toast.net show both networks being awfully
fast... but pulling from Sourceforge.net etc don't. Must be gaming the
tests.

I'm considering getting a T-1. I worked in an ISP in 1998 and sat on a T-1
and could tell when pages were being pulled from disk. Would getting a T-1
help or are they also run through some crappy network like TimeWarner and
BellSouth seem to be?  I've run a few tracert's and see high times. Any
sugestions?

TIA,
TimJowers
PS> tracerts. 2-5 are TWX network from what I can tell. Seems to be where
the crap is in the network. But why does #11 suddenly have a good time?
[root at localhost ~]# tracert www.kernel.org
traceroute to www.kernel.org (204.152.191.37), 30 hops max, 40 byte packets
 1  192.168.1.1 (192.168.1.1)  1.642 ms  1.700 ms *
 2  * * *
 3  * * *
 4  * * *
 5  * * *
 6  * pos14-0.rlghncrdc-rtr1.nc.rr.com (24.25.0.5)  634.242 ms  634.387 ms
 7  son0-0-3.rlghncrdc-rtr3.nc.rr.com (24.93.64.34)  634.546 ms  634.650 ms
634.801 ms
 8  te-1-3.car1.Raleigh1.Level3.net (4.71.160.5)  635.320 ms
te-1-4.car1.Raleigh1.Level3.net (4.71.160.1)  635.447 ms  635.564 ms
 9  ae-11-11.car2.Raleigh1.Level3.net (4.69.132.174)  643.455 ms  643.597ms
643.712 ms
10  ae-6-6.ebr2.Washington1.Level3.net (4.69.132.178)  649.176 ms  648.898ms
649.050 ms
11  * * ae-2.ebr2.Chicago1.Level3.net (4.69.132.69)  254.990 ms
12  ae-1-100.ebr1.Chicago1.Level3.net (4.69.132.41)  248.797 ms * *
13  ae-3.ebr2.Denver1.Level3.net (4.69.132.61)  234.408 ms  423.459 ms
533.733 ms
14  ae-1-100.ebr1.Denver1.Level3.net (4.69.132.37)  520.128 ms  622.826 ms
623.019 ms
15  ae-3.ebr2.SanJose1.Level3.net (4.69.132.57)  651.141 ms  651.347 ms
651.529 ms
16  ae-2-56.bbr2.SanJose1.Level3.net (4.68.123.161)  747.479 ms  747.158 ms
747.371 ms
17  so-13-0.hsa3.SanJose1.Level3.net (4.68.121.234)  750.664 ms  750.859 ms
751.089 ms
18  isc-level3-ge.SanJose1.Level3.net (4.68.111.59)  751.637 ms  703.822 ms
703.706 ms
19  zeus-pub2.kernel.org (204.152.191.37)  742.215 ms  741.886 ms  741.545ms
[root at localhost ~]#
1 second latency!
Looks like zeus is on Level3's network. (Level3 bought ACSI from what I
remember.)
[root at localhost ~]# tracert www.microsoft.com
traceroute to www.microsoft.com (207.46.19.60), 30 hops max, 40 byte packets
 1  192.168.1.1 (192.168.1.1)  1.714 ms  1.771 ms  1.844 ms
 2  10.127.0.1 (10.127.0.1)  480.874 ms  481.037 ms  481.109 ms
 3  gig2-1.rlghncj-rtr3.nc.rr.com (24.25.1.18)  481.181 ms  481.323 ms
481.395 ms
 4  srp5-0.rlghncg-rtr2.nc.rr.com (24.25.2.147)  482.724 ms  482.796 ms
482.869 ms
 5  pos1-1.rlghnca-rtr2.nc.rr.com (24.25.20.25)  482.940 ms  483.014 ms
483.086 ms
 6  pos14-0.rlghncrdc-rtr2.nc.rr.com (24.25.0.9)  490.783 ms  368.689 ms
368.823 ms
 7  son0-1-1.chrlncsa-rtr6.carolina.rr.com (24.93.64.57)  373.686 ms
317.454 ms  317.552 ms
 8  te-1-3.car1.Charlotte1.Level3.net (4.71.124.1)  336.068 ms  364.222 ms
te-1-4.car1.Charlotte1.Level3.net (4.71.124.5)  364.848 ms
 9  ae-4-4.ebr1.Atlanta2.Level3.net (4.69.132.162)  405.642 ms  369.521 ms
370.828 ms
10  ae-3.ebr1.Dallas1.Level3.net (4.69.132.81)  397.114 ms  404.205 ms
404.345 ms
11  ae-1-100.ebr2.Dallas1.Level3.net (4.69.132.46)  392.582 ms  426.240 ms
426.348 ms
12  ae-3.ebr2.LosAngeles1.Level3.net (4.69.132.77)  450.449 ms  446.432 ms
446.546 ms
13  ae-1-100.ebr1.LosAngeles1.Level3.net (4.69.132.5)  447.036 ms  515.681ms
515.840 ms
14  ae-2.ebr1.SanJose1.Level3.net (4.69.132.9)  413.698 ms  280.643 ms
280.442 ms
15  ge-2-0-0-53.gar1.SanJose1.Level3.net (4.68.123.66)  287.572 ms  302.898ms
303.002 ms
16  * * *
17  * * *
18  * * *
19  * * *
20  * * *
21  * * *
22  * * *
23  * * *
24  * * *
25  * * *
26  * * *
27  * * *
28  * * *
29  * * *
30  * * *
Looks like a bunch o routers between Level3 and Microsoft's public server.
I'd hate to see that network diagram.
Now this is what one would expect!!!
[root at localhost ~]# tracert www.sourceforge.net
traceroute to www.sourceforge.net (66.35.250.203), 30 hops max, 40 byte
packets
 1  192.168.1.1 (192.168.1.1)  0.457 ms  0.540 ms  2.150 ms
 2  10.127.0.1 (10.127.0.1)  11.696 ms  11.888 ms  19.011 ms
 3  gig2-1.rlghncj-rtr3.nc.rr.com (24.25.1.18)  23.058 ms  26.573 ms  30.432ms
 4  srp5-0.rlghncg-rtr1.nc.rr.com (24.25.2.148)  35.097 ms  40.161 ms
44.200 ms
 5  pos1-0.rlghnca-rtr1.nc.rr.com (24.25.20.17)  44.602 ms  45.033 ms
45.161 ms
 6  pos14-0.rlghncrdc-rtr1.nc.rr.com (24.25.0.5)  44.420 ms  11.120 ms
14.739 ms
 7  son1-0-1.rlghncrdc-rtr3.nc.rr.com (24.93.64.9)  19.538 ms  7.590 ms
12.022 ms
 8  te-1-3.car1.Raleigh1.Level3.net (4.71.160.5)  15.428 ms
te-1-4.car1.Raleigh1.Level3.net (4.71.160.1)  9.469 ms  13.685 ms
 9  ae-11-11.car2.Raleigh1.Level3.net (4.69.132.174)  18.685 ms  7.515 ms
12.209 ms
10  ae-6-6.ebr2.Washington1.Level3.net (4.69.132.178)  33.948 ms  34.071 ms
13.564 ms
11  ae-23-52.car3.Washington1.Level3.net (4.68.121.48)  63.807 ms * *
12  savvis-level3-te.Washington1.Level3.net (4.68.110.102)  23.938 ms
15.046 ms  20.213 ms
13  cpr2-ge-5-0.VirginiaEquinix.savvis.net (204.70.193.101)  24.320 ms
14.638 ms  15.086 ms
14  bcs2-so-2-0-0.washington.savvis.net (204.70.193.153)  21.844 ms  19.370ms
22.427 ms
15  bcs1-so-7-0-0.Washington.savvis.net (204.70.192.33)  29.405 ms  18.626ms
21.358 ms
16  dcr1-so-0-3-0.Chicago.savvis.net (204.70.192.26)  40.107 ms  44.019 ms
47.662 ms
17  dcr2-so-5-0-0.Chicago.savvis.net (204.70.192.46)  66.482 ms  67.376 ms
32.717 ms
18  dcr1-so-4-3-0.Denver.savvis.net (204.70.193.225)  61.265 ms  64.227 ms
69.515 ms
19  dcr2-so-1-1-0.SanFranciscosfo.savvis.net (204.70.193.218)  107.809 ms
115.842 ms  123.860 ms
20  bhr1-pos-0-0.SantaClarasc8.savvis.net (208.172.156.198)  124.176 ms
128.262 ms  128.444 ms
21  csr2-ve243.santaclarasc8.savvis.net (66.35.194.51)  129.681 ms  129.863ms
151.883 ms
22  66.35.210.202 (66.35.210.202)  129.258 ms  96.820 ms  90.873 ms
23  sourceforge.net (66.35.250.203)  91.554 ms  92.252 ms  91.373 ms
I guess sitting idle for 5-10 minutes might be the trick? Maybe they have
some throughput throttling or something on the network? ????? Anyone know?
Or maybe they just have limited bandwidth in their TWX datacenters so while
one could theoretically get 1.5Mb/s in reality they have 100,000 people all
on one 10Mb ethernet?  :-) Seems like it.