I passed this exam first time. The way I did it was to use the question testing software at the back of the book and kept doing all tghe questions in every chaper/section until I got them all right. You could try the same with a transcender exam.
Pay special attention to the Package and Deployment wizard and read about the way Microsoft expects things to be done. This comment rrelates to all the MCSD exams. I disagree with a lot of what Microsoft says, but they want you to answer there way in the exam.
The exam askes a lot of scenario questions. The trick here si to decode the scenario into basic, easily understood requirements. The question is the easy to tackle.
A good example in my exam was a queation about the best way to build a system. It was a finance application that accessed a legacy database/system aswell as a new one. Performance was important, but so was reliability and availability, as well as easily maintainable. The legacy system was flakey and crashed often, and would bring down any program accessing it. The question was what would you build. The choices were a single application, and application with a DLL, an application with an ActiveX EXE, or an application with a Custom OCX to contain the legacy system logic/code.
The answer was simple when you split the question up. Perfomance would point to the Single EXE, or the DLL, or the OCX as these perform better than the Application with an ActiveX EXE. However, availability and reliability were important. If any of the last three were used and the Legacy system went down then the whole application would go down as the DLL and OCX run in process. The ActiveX EXE runs out of process so when it goes down it won't bring the whole application down. That is the choice Microsoft wants.
You need to understand the various architectures and what the benefits/side effects of each one is. You then need to practice decoding the questions and mapping them to what you know.
James

James Culshaw
jculshaw@active-data-solutions.co.uk