Did you send those messages concurrently, or one after the other? If you care about scalability, simple sequential tests aren't going to tell you much. You'll want to construct a test where you measure how the response time varies as you increase the load, preferably from a number of clients.
But then you say that you're interested in all processes being on the same machine, which implies that you don't in fact care about scaling beyond, well, one. So what exactly are your requirements when you say you are interested in "performance and scalability"? What characteristics do you expect the inter process communication in the application you're planning to write to have? -- Ian Griffiths DevelopMentor -----Original Message----- From: wycklk [mailto:[EMAIL PROTECTED] I wrote 2 processes running on the same machine. Each process communicates with each other by sending 200 messages with a size of 1024 bytes. I got the following result for each IPC methods: TCP Socket: 10ms Remoting: 100ms (TCP channel, binary format serialization) System.Messaging: 40ms (Express mode, XML format serialization) System.Messaging: 20ms (Express mode, Binary format serialization) The purpose of this experiment is to find out what is best IPC methods in terms of performance and scalability, if all my processes run on the same machine. For remoting, I use Singleton object call to pass the 1024 bytes data. For System.Messaging, I send and receive 100 messages from the same program. Anybody to share similar experience ?
