Client Server and ADO Performance Test What is really being tested?
I am evaluating Segue's SilkPerformer for our client server application which uses ADO to communicate with the database. The SilkPerformer script mimics the ADO calls to the database. So, my question is, am I really testing the performance of the client server app. or the performance of the database? BTW, Seque has been very helpful in working through issues and the tool is one of the better ones I have seen for Performance testing.
Re: Client Server and ADO Performance Test What is really being tested?
What you are really testing is the server side of the c/s application. The numbers you will get back are the actual time from when the packet leaves the client's network card until the response gets back to the client's network card.
I haven't started using Silk yet (I actually start next week) so someone correct me if I'm not 100% on about this next part...
I believe Silk will then add in either a static or variable ammount of time to SIMULATE the client side processing time. This amount of time generally represents the time the client side processing took when you recorded the script. This time should be completely editable by you.
Sr. Performance Engineer