This week's giveaway is in the EJB and other Java EE Technologies forum. We're giving away four copies of EJB 3 in Action and have Debu Panda, Reza Rahman, Ryan Cuprak, and Michael Remijan on-line! See this thread for details.
Hi, I am trying to improve the performance of a particular bean that accesses the database something like addCustomer(customerVo). I have written a imporved version of this. I call the imporved version and clock the time in a jsp. I also call the older version and clock the time in the same jsp. The problem is that sometimes the new version of the bean performs better and sometimes it does not perform better. How do I decide what is to be done. Do I accept the changes or dont I?? I am working on Weblogic 6.1 on a Windows 2000 machine with jdk 1.3 installed. To Be or Not to Be is the Question - Perhaps By William Sheakspere. [ December 19, 2002: Message edited by: Rahul Mahindrakar ]
Tough to say. My first guess would be that other processes are interacting with either the database, or the networkm, or the CPU that your program is running on. Ideally I'd just keep testing, paying attention to any measurements of DB usage, network traffic, or CPU usage that I could get my hand on until I had a better idea what the problem was. Failing that, I'd try to just test enough to get a sense of the "average" behavior. Is the new program faster more often than it's slower, or the other way around. If ultimately you can't really tell the difference, then I guess there's no real incentive to make the change. Then too, there's no real reason not to, either, provided you're confident in your testing process to know that the new code doesn't introduce new bugs. That's a pretty wishy-washy answer I know, but it's the best I have. Anyone else?