New results: I realised that if i reduce the number of flows to 3 instead of 5, tcpdump will show graphs which will corresponds to that of TTT. Anyone has any idea why?? I doubt it is the snaplen. ----- Original Message ----- From: "Rick Goh" <rickgoh@xxxxxxxxxxxxxxx> To: "lartc" <lartc@xxxxxxxxxxxxxxx> Sent: Friday, March 23, 2001 5:07 AM Subject: [LARTC] Tcpdump capturing faulty??? > Hello, > I am testing traffic control on linux. I used Tele-traffic tapper > (http://www.csl.sony.co.jp/person/kjc/programs.html) to see real-time > throughput of my client and it shows that traffic control of different type > of traffic flow works! However when i used tcpdump, it shows that priority > doesn't work, meaning all flows have the same bandwidth => tc doesn't work. > > In detail: > > Experiment 1: With borrow, FAILED. With all streams bounded, works. > When i set on 192.168.1.10: > port 42011 --> 5 (Mbps) > port 42012 --> 10 > port 42013 --> 15 > port 42014 --> 20 > port 42015 --> 50 > > Experiment 2: > When i set on 192.168.1.10: > port 42011 --> Priority 1 > port 42012 --> Priority 2 > port 42013 --> Priority 3 > port 42014 --> Priority 4 > port 42015 --> Priority 5 > > > On Tele traffic tapper (TTT), works fine but not tcpdump. What do i mean by > tcpdump cannot work: > 1. After i tcpdump it, TTT can actually read from the dump file. But it > shows Traffic control doesn't work; i.e. Meaning all 5 streams share the > same bandwidth. > > 2. I tcpdump into a file and post analysis it by breaking it into the > traffic from each different port. And i graph it. DOesn't work!! Meaning all > 5 streams share the same bandwidth. > > Has anyone come across this major hiccup??? > > > Regards. > > > > > > _______________________________________________ > LARTC mailing list / LARTC@xxxxxxxxxxxxxxx > http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://ds9a.nl/2.4Routing/
