​netperf

The netperf tool is strong choice for both throughput and latency testing.

The TCP_​STREAM test in netperf measures the throughput capabilities of the device.

Server command example:

Client command example: netperf -t TCP_STREAM -l 30 -H <serverIP>

The TCP_​RR test in netperf returns latency in a value of transactions/second. This is a round-trip number. It is recommended to use the -T x,x option, were “x” is CPU local to the device.

The one-way latency can be calculated using: Latency(usec)=(1⁄2)/[Transactions/sec]*1,000,\

Server command example: netserver

Client command example: netperf -t TCP_RR -l 30 -H <serverIP> -T x,x

To start multiple instances (threads) of netperf, Intel recommends using a for-loop to map threads to TCP ports and running netperf in the background using & to create multiple processes in parallel.

Server command example, start 8 threads: port=""; for i in {0..7}; do port=520$i; bash -c "netserver -L $serverIP -p $port &"; done;

Client command example, start 8 threads: port=""; for i in {0..7}; do port=520$i; bash -c "netperf -H $serverIP -p $port -t TCP_STREAM -l 30 &"; done;