Chapter 20: Performance Tuning ATL Server Web Applications


ATL SERVER PROVIDES AN efficient, high-performance framework for writing Web applications and Web services. However, ATL Server can t provide facilities for every scenario encountered by users, and it can t guard against every programming idiom that may hinder performance.

In this chapter we discuss techniques that will help you write high-performance code within the ATL Server framework. We examine how you can leverage portions of the ATL Server library in writing new user code and how to best use the services provided by ATL Server (such as caching, dynamic services, and per-thread services) to write Web applications that perform well. Additionally, we look at performance pitfalls, such as specific programming techniques to avoid when writing high-performance Web applications. Finally, we examine methods of accurately measuring the performance of Web applications and diagnosing problem areas: which metrics are relevant, testing tools and hardware, and testing setup.

What Performance Is

When dealing with definitions of performance, it s important to keep in mind some important caveats. First, it s practically impossible to duplicate the same environment in a lab performance-testing setting that users of an application will encounter; therefore, whatever performance numbers come from a lab setting won t necessarily reflect the performance users can expect. However, these kinds of performance numbers still serve as the best guide in improving your application s performance.

Second, even if it were possible to faithfully duplicate an end-user environment in a laboratory setting, the specific performance metrics you choose to measure may not necessarily be the same metrics that would be most relevant in improving performance from an end-user perspective.

And the third and most important caveat is this: The information about performance in this chapter is provided as general performance guidelines, suggestions, and testing scenarios ”no general framework can accurately reflect all the performance- related considerations for specific applications. Performance testing and performance tuning must be considered in relation to the specific application.

Having put forth all these caveats, we feel it s important to note that there are still many useful generalizations that can be made about performance testing and performance tuning that will apply to most applications. The information in this chapter is still useful and accurate for the majority of applications built with ATL Server. It s just important to be able to recognize when it isn t.




ATL Server. High Performance C++ on. NET
Observing the User Experience: A Practitioners Guide to User Research
ISBN: B006Z372QQ
EAN: 2147483647
Year: 2002
Pages: 181

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net