Not sure what to say or what your goals are for this paper..
This paper may be useful as a guideline for *your* specific test team, but it makes a LOT of assumptions that are not applicable to all "performance testing" projects.. and then these assumptions are never stated.
You jump right into talking about databases and application servers and web servers.. What kind of system is this? What if my system doesn't have a database? or isn't web based? Do I really need to prepare SQL scripts if we don't have a database? You talk about setting up your Rational tools.. what if I don't use Rational tools?
I think there may be some good information in there if you presented it as more of a case study. You could begin by talking about the environment.. the type system.. challenges.. constraints.. etc.. then go through your test setup.. descriptions.. guidelines... etc.. and finally, draw some conclusions.
If however, this is just a presentation of "performance testing" in general, I think it is severely lacking.
I'm with Corey. This looks more like a test strategy or test plan document for a specific project at a specific company, versus a generic white paper. I have tested applications that have no databases attached.
One thing jumped out at me right away. You title the paper "Performance Testing", and then state that "performance test activity of any application aims at finding the response time of the identified business functions at defined user load with expected record volume in the application database." That pretty much corresponds with the classic definition of "Load Test", which is but one of several types of preformance tests. What if my performance test aimed at finding out the maximum traffic that a server could handle? That does not fit your definition of performance test. What if I am testing my app without a database? That doesn't fit your definition either. How about stability tests? The point is that there are many types of performance tests, and your paper makes it look like there is only one.
You also do not address hardware in your definition, which is fully one half of the performance testing equation.
Your paper is coercing the specific into the generic, which is not generally a good thing. In this case, you cannot make generic statements about performance testing and how to do it based on a single instance or type of test, and have the paper stand up to scrutiny.
However, instead of White Paper, it could have been referred as a checklist for applications having 2 or 3 tier architecture.
<font size="2" face="Verdana, Arial, Helvetica">It still would have been way off. It could be referred to as a checklist for a specific project you worked on that is generally not applicable to any other situation.