[Tristaan]

Hi.

We're using an Indy 9 Server component to handle requests in one of our applications. In a single user environment (1 user at a time sending requests) everything is performing well with approximately 2-3 seconds elapsiong for each request.

However, when I start increasing the number of simultaneous users (2,5, 10, 20, etc), with each increase, I see a major decrease in performance (around 16 seconds a request for 10 users).

I'm using TestComplete 3 by AutomatedQA Corp to perform my load tests. There are two statistics logged for each user request that I'm using to guage performance. One is the Reponse Time which is the amount of time the server takes to process the request and the data transmission time of the response. Consistantly, no matter how many users, this is no more than 0.4 seconds.

The one that we are concerned about is the reqest time which is the amount of time it takes for the user request to get to the server. It is this time that constistantly increases with the more users I "throw" at the server.

We're a little dumbfounded at how there could be such a major difference in performance in supposedly a multi-threaded component designed to handle web traffic. Anybody up here familiar enough with the Indy servers and with http communications to help us understand why this is happening?

Thanks!