For the measurement of the latency, throughput, I'm using the "Example Performance Test for RTI Data Distribution Service (updated 07/23/10)" given in the "Performance Section" of the "Public Knowledge Base"
 
There is some confusion regarding the interpretation of the results. The Subscriber prints out the "Throughput" results and one of the paramters it prints out is "Packets/sec (Also Packets/sec {average}). I understand one can control the message size one wishes to send (from 32 bytes to 63, 000 bytes) by using the "-datalen" command line option. Now, does "Packet/sec" as printed out on the subscriber screen represent just the data or the data and the overhead ?
 
My problem is such; I want to send a 200 byte message 10, 000 times a second, 20, 000 times a second, 30, 000 times ..... 100, 000 times a second and measure the corresponding throughput/latency. My other question is how do I configure the number of times a second I want to send this 200 byte message a second.
 
I read the corresponding documentation that comes with this program and to the best of my knowledge, I need to do the following but I am not sure.
 
Publisher;
 
perftest_cs -pub -datalen 200 -spin<>
 
Subscriber;
 
perftest_cs -sub