Learn the proven performance tuning methodology that I used to successfully tune the performance of Enterprise Java environments for companies ranging from small 10-person development teams to large-scale Fortune 500 companies.
Last week I released a new white paper entitled, A Formal Performance Tuning Methodology: Wait-based Tuning
, that I'd like to share with you. If you are a regular reader you'll know that while I'm focused on designing performance management software now, I spent a couple years on the road resolving performance problems for Quest Software's customers. During this time I built and validated a performance tuning methodology that I call wait-based tuning. In this methodology, rather than looking at an application and an application server from a purely metrics perspective (e.g. what is the thread pool usage, how many statements in a prepared statement cache are being removed, and so forth) I instead look at tuning an application from an application architecture perspective. I ask the question: in my application architecture, what application server services (thread pools, database connection pools, caches, pools, tier separations, and so forth) can cause my application to wait?
After an architecture analysis I have a list of "wait points" that I can use to more effectively tune my environment. While this methodology is applicable to tuning a production environment, this paper is written from a performance testing perspective. By implementing this tuning methodology in a testing initiative, you can discover problems that will eventually affect your application in a production environment when your application is subjected to high amounts of load. This is something that I really recommend you do before you ever launch an application in a production setting - find your problems now, before your users find them!
And if you're interested in performance testing and performance tuning, join me tomorrow (March 18th) as I start a new three part web cast series called the Java Performance Winner's Circle. Tomorrow's web cast is entitled Best Practices in Java Environment Testing
, through which I will describe when and how to implement performance testing in your application development life cycle. The two follow up web casts are: "Wait-based Tuning for Enterprise Java Environments", which reviews the contents of the aforementioned white paper, and "Achieving Continuous Performance Management", which discusses the best practices for implementing automated performance tests in a continuous integration environment. I hope to see you at all of these (and if you attend at least two of the web casts, Quest Software will give you a "Java Performance Winner's Circle" T-Shirt.)
Send your comments about the white paper and web cast series to me at firstname.lastname@example.org.