aspose file tools*
The moose likes Performance and the fly likes How to Decide what to do Big Moose Saloon
  Search | Java FAQ | Recent Topics | Flagged Topics | Hot Topics | Zero Replies
Register / Login
JavaRanch » Java Forums » Java » Performance
Bookmark "How to Decide what to do" Watch "How to Decide what to do" New topic
Author

How to Decide what to do

Rahul Mahindrakar
Ranch Hand

Joined: Jul 28, 2000
Posts: 1859
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 ]
Jim Yingst
Wanderer
Sheriff

Joined: Jan 30, 2000
Posts: 18671
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?


"I'm not back." - Bill Harding, Twister
 
 
subject: How to Decide what to do