[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