You work as the Enterprise application developer at Domain.com. The Domain.com network consists of a single Active Directory domain named Domain.com. All servers in the domain run Windows Server 2003. Your responsibilities at Domain.com include the support and deployment of applications. You completed the development of a Microsoft ASP.NET Web application using Microsoft Visual Studio 2005 Team Edition for Software Developers. You then deploy the Microsoft ASP.NET Web application to a production server. Then you discover that the central processing unit (CPU) usage on the Web server sometimes reaches 100%. You then assume that the Web application can be the cause of the sudden increase of CPU usage, but you have no idea of where in the application the problem can be.
You now need to ascertain the reason why this performance spike occurs.
What should you do?
A.
You should create a Web test that consists of unit tests and run the application.
B.
You should create a Web test that consists of a load test and run the application.
C.
You should create a Performance test in Visual Studio and instrument the application.
D.
You should create a Performance test in Visual Studio and sample the application.
Explanation:
Sampling periodically interrupts the application to collect performance data. This is extremely helpful in cases where you are unsure of where in an application there might be a bottleneck.
Incorrect answers:
A: You should not create a Web test to test a Web application’s performance. You know that the application performs, what you need to do is to locate the performance bottleneck.
B: A Web test consisting of a load test will allow you to test whether the application performs as expected when a certain number of users are accessing the application. In this case it is known that the application does perform, what is not known is where the performance bottleneck exists.
C: Instrumentation presupposes the location of the performance bottleneck is known and in this case you are unsure of where in the application the performance bottleneck is occurring.