[TriLUG] really slow web. help!

P. Steven Lorenz pstevenlorenz at gmail.com
Mon Mar 26 16:38:56 EDT 2007


I am in Cary using TWC as well.  This might help a bit.
I had similar problems and switched to OpenDNS that helped a little bit.
 I also modified some stuff in Firefox and that things up considerably.

OpenDNS might help as it dropped my latency to about 1/10th of what is
was.

Steven
Tim Jowers wrote:
> 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.




More information about the TriLUG mailing list