Hi Alex,
Thank your for your reply, I got your point, I have to combine business needs and operational environment as you said.
I got your idea, Idea is simple but the implementation is not.
Right now business need is like this, my customer wants to learn how many user can connect and use their system concurrently. So, this is the requirement.
And for the environment lets say that I have a fixed environment, hardware and code, and I want to learn max. concurrent users for the current system with current configuration, (not ideal max users for the system), how will I detect max.users ?
As you wrote, In my tests, I use performance counters within Run-Time graph, from server.
During a test, I got 500 internal server error and see that some of vusers could not login.
This warning are returned bcos of http protocol standarts right ?
I wonder , why they could not connect ?Maybe if the system can put them in a queue, and make connection later , when the possible user login, then make other connections, ?
Is there a way within TestComplete to extend response times ? Where can I see max response time for a request ?
Under tools, options, http load , there is receive and send time -out limits and both of them are 600s by default.
I think this is not request time bcos it is too much for a request.
I real users waits some seconds more for website response, is that crash ? How can calculate users for a system ?
Thank you.
