I want to do some performance evaluation of my cache algorithm to see how it performs when doing read and update. What would be the best way to do that when multiple client want to read and update from same server node. So far creating 10 threads for a client and then dong remote command can work but it will count the time for JNDI lookup and RMI. Also thread are useful when we working in same jvm and same process. While I think server runs in separate JVM and is separate process. Also is there any way to run different set of clients from within each Application Server(container) with some little time variation say that I can sea how the performance of my algorithm running on a cluster.
I have 8 nodes in a cluster.
I want to do read and update 10 clients doing read and write on a node
As there are 8 nodes so I want to 80 clients (10 each server) doing read and write through out server.
How can I achieve it? through stand alone app client and threads or there is some other way? while searching I came across cone unix command to start clients simultaneously or within some time. What other option do I have to do so and which one is the best.