Rainer Duffner wrote: > srv3# mtr -c 10 -r centos.org > HOST: srv3.ultra-secure.de Loss% Snt Last Avg Best Wrst StDev > ... > 7. ae-92-92.csw4.Washington1.Le 0.0% 10 98.6 102.3 97.0 110.3 5.0 > 8. ae-91-91.ebr1.Washington1.Le 50.0% 10 101.8 103.6 98.6 109.8 4.1 > 9. ae-2.ebr3.Atlanta2.Level3.ne 30.0% 10 115.2 118.9 114.1 124.7 4.4 > 10. ae-7.ebr3.Dallas1.Level3.net 20.0% 10 132.4 135.5 130.9 142.2 4.7 > 11. ae-73-73.csw2.Dallas1.Level3 0.0% 10 143.9 137.7 131.2 143.9 4.9 > ... > So, from this point of view, L3 seems to be losing some packets. > 'lost packets' on traces from intermediate hops aren't significant, as most routers treat PING as a low priority and if they are busy doing their primary routing job, they'll ignore some percentage of pings. traceroute and its ilk are very inexact tools. your MTR command, run from a San Jose, California machine, gives... # mtr -c 10 -r centos.org XXXXXXXXX.com Snt: 10 Loss% Last Avg Best Wrst StDev ge-0-0-1-414.er2.sjc1.got.net 0.0% 2.5 22.0 2.5 191.4 59.5 ge-1-0-10.cr1.sjc1.got.net 0.0% 22.2 22.2 14.5 30.0 5.6 173.ge-1-3-0.mpr1.sjc2.us.above.net 0.0% 3.9 12.0 3.0 48.5 17.0 xe-1-1-0.er1.sjc7.us.above.net 0.0% 5.5 4.9 3.4 7.0 1.2 above-level3.sjc7.us.above.net 0.0% 66.3 11.1 3.2 66.3 19.5 vlan69.csw1.SanJose1.Level3.net 0.0% 7.6 9.7 4.5 17.0 4.2 ae-63-63.ebr3.SanJose1.Level3.net 0.0% 17.8 13.8 6.5 30.4 7.6 ae-2.ebr3.LosAngeles1.Level3.net 0.0% 23.6 18.1 12.5 25.4 5.2 ae-83-83.csw3.LosAngeles1.Level3.net 0.0% 24.4 20.3 13.3 25.3 4.9 ae-82-82.ebr2.LosAngeles1.Level3.net 10.0% 24.8 21.8 14.9 26.2 4.1 ae-3.ebr3.Dallas1.Level3.net 20.0% 47.6 48.8 46.7 56.1 3.0 ae-73-73.csw2.Dallas1.Level3.net 0.0% 46.1 50.3 46.1 58.4 4.3 ae-22-79.car2.Dallas1.Level3.net 0.0% 45.9 47.0 45.9 48.0 0.8 DATABANK-HO.car2.Dallas1.Level3.net 0.0% 47.6 47.4 46.2 49.0 1.0 ae_cw_10g.databank.com 0.0% 48.5 47.1 46.7 48.5 0.6 pod22c_ae.layeredtech.com 0.0% 46.9 47.5 46.5 48.2 0.5 ??? 100.0 0.0 0.0 0.0 0.0 0.0 162.194.232.72.static.reverse.ltdomains.com 0.0% 48.0 47.3 46.4 48.0 0.4