IEPM

Bulk Throughput Measurements - RIKEN, Japan


Bulk Throughput Measurements | Bulk Throughput Simulation | Windows vs. streams | Effect of load on RTT and loss | Bulk file transfer measurements

SLAC to RIKEN

On November 30, 2001, measurements were made between from pharlap.slac.stanford.edu to afs0.riken.go.jp  Afs0 was a PIII *497MHz processor running LInux 2.2 with a 100Mbps NIC located at the RIKEN Lab in Japan. Pharlap was a Sun E4500 with 6*336MHz cpus and a GE interface running Solaris 5.8. SLAC had a 1Gbps link to the Stanford campus and from there a 622Mbps link to CalREN and Internet 2. The routes  in both directions use Internet 2. The ping response for 226 default length (64 Bytes) from SLAC to RIKEN was min/avg/max (std) = 245/336/3624 (std = 454). The pipechar from SLAC to Rice was also recorded. The window buffer sizes on pharlap are shown below:
ndd /dev/tcp tcp_max_buf = 4194304
;ndd /dev/tcp tcp_cwnd_max = 2097152
;ndd /dev/tcp tcp_xmit_hiwat = 16384
;ndd /dev/tcp tcp_recv_hiwat = 24576

The window buffer sizes on afs0 are shown below:
more /proc/sys/net/core/wmem_max = 8388608
;more /proc/sys/net/core/rmem_max = 8388608
;more /proc/sys/net/core/rmem_default = 65536
;more /proc/sys/net/core/wmem_default = 65536

The performances shows that the maxima (the top 10% measurements) achieve over 15.7 Mbits/s throughput. To achieve a maximum one needs at least 25 to 30 streams and a window size of at least 64KBytes.. 


Created November 30, 2001, last update November 30, 2001.
Comments to iepm-l@slac.stanford.edu