IEPM

Bulk Thruput Measurements-LBL

SLAC Home Page
Bulk Thruput Measurements | Bulk throughput simulation | Windows vs. streams | Effect of load on RTT and loss | Bulk file transfer measurements

LBL to SLAC

The following graphs shows the thruput measured from LBL, Berkeley California (costard.lbl.gov a Debian Linux machine) to SLAC (pharlap.slac.stanford.edu) on 10/23/00. At this time the route was going via Chicago, the actual route can be seen by looking at the pchar output. The window sizes choosable were limited to less than 128kBytes by costard. The maximum thruput is about 56Mbits/sec. The best thruput with the minimum window size I was able to choose (16kBytes) was about 10Mbits/s. The best thruput with 1 stream (128kByte window) was about 6Mbits/s. With the default of 1 stream and 16kByte window the thruput was about 560kbits/sec. The thruput seems to saturate at about 9 streams with 128kByte windows at about 54Mbits/s. The default window size on costard was 64kBytes and with 1 streams and 64kB window we measured 1.6Mbits/sec.
LBL to SLAC bulk thruput contour map LBL to SLAC saturation

Sample Bulk Throughput Measurements

Cal Tech | CERN | Colorado | IN2P3 | INFN |Daresbury | Stanford to Daresbury | ANL |BNL | LBL |SLAC To Stanford's Campus

BACK TO MAIN PAGE

Summary of Bulk Throughput Measurements

Created August 25, 2000, last update August 29, 2001.
Comments to iepm-l@slac.stanford.edu