To change your JDJList options, please visit: http://www.sys-con.com/java/list.cfm-----Original Message-----
From: Greg Nudelman [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, August 27, 2002 4:04 PM
To: JDJList
Subject: [jdjlist] RE: benchmark programDear Jeff,We have *almost* the same architecture here... we actually have many of the same Applet problems. Typically we just check the configuration verbally with the client using standard PC utilities and then turn on the system trace to check how long the system takes to respond. The key is to set the "user system minimum hardware requirements" high, then you will have fewer problems with users calling in. Unfortunately, not everyone can afford to change their whole office to IE 5.5 SP2+ 1000MHz P3s with 256K RAM. :-(I would also be interested in something automatic like that. However, I would suspect you may need some kind of signed applet to do this, as some of your measurements may violate the Applet sandbox model.Let us know what you'll find!GregP.S. I found a decent article on some Applet testing issues that may be helpful:-----Original Message-----
From: Yan, Hong [IT] [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, August 27, 2002 12:42 PM
To: JDJList
Subject: [jdjlist] RE: benchmark programTo change your JDJList options, please visit: http://www.sys-con.com/java/list.cfmGreg:Thank you for suggestion. Actually, the program that I need is a program that benchmarks remote user's JVM.We have a Applet / J2EE based transaction system, and I need to write an Applet for the user to run and benchmarks user's environment ( - so that when a user complains that the program is slow, our tech support guys can instruct the user to run this program and find out how good the local operation environment is and how good the internet transmission is).I looked into JUnit, and I found that it is basically a testing platform, and the Catus/J is interesting but unfortunately it is not for benchmarking the environment.Any suggestions?Jeff-----Original Message-----
From: Greg Nudelman [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, August 27, 2002 3:25 PM
To: JDJList
Subject: [jdjlist] RE: benchmark programTo change your JDJList options, please visit: http://www.sys-con.com/java/list.cfmCheck out Cactus and JUnit -- these are the open source Java testing frameworks that may be of help. But I think you may need to write the code you want to test yourself. After all, who really cares about some else's app? It is _your_ app that you want to benchmark, right?
Cactus and JUnit should help there.
Greg
-----Original Message-----
From: Yan, Hong [IT] [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, August 27, 2002 12:09 PM
To: JDJList
Subject: [jdjlist] benchmark program
Hi all:
I am looking for a benchmark program (java class) to test different versions
of JVMs.This program has to be small and can be wrapped in an applet.
I would like some suggestions. Thanks
Jeff
To change your JDJList options, please visit: http://www.sys-con.com/java/list.cfm
To change your JDJList options, please visit: http://www.sys-con.com/java/list.cfm
Title: RE: [jdjlist] benchmark program
Greg:
Thanks
for the info. I found the article you mentioned very
interesting.
Signing an applet is not a big deal to us, because we have all the other
(transactional) applets signed. But I do want to try a sandbox algorithm first,
if possible at all.
The
benchmark that we want is "relative", in other words, not based on hardware
configurations such as PENTIUM xx, RAM xxx, etc. We want to run a series of
tests, including local computations and cross-internet data transfers and then
display a report in the browser to remind user that something is lower than what
we expect.
However, we are still in a research stage...it is difficult to define
what is "what we expect". There could be users using a dial-up, and users using
a T3 line, etc.
Currently, I find JBench very interesting...It has a framework of timing
threads. (see http://www.yoda.arachsys.com/java/jbench/docs/index.html)
rgds
Jeff
- [jdjlist] RE: benchmark program Greg Nudelman
- [jdjlist] RE: benchmark program Yan, Hong [IT]
- [jdjlist] RE: benchmark program Greg Nudelman
- [jdjlist] RE: benchmark program James A. N. Stauffer
- [jdjlist] RE: benchmark program Yan, Hong [IT]
- [jdjlist] RE: benchmark program Yan, Hong [IT]
