Yeah, I totally agree with you.

My main point that I wanted to make was that the kernel configuration could have drastic effects on the system performance. And also that RH has many different "stock" kernels that could be installed. Their "enterprise" kernel has pretty different parameters compiled in than the "regular" kernel. I've also seen linux boxes where the installer didn't properly detect SMP and installed a kernel built for one processor only.


Dan Christopherson wrote:

You're comparing apples with oranges, bananas with baboons. If your dual processor box has 'slow' drives and your laptop's IDE has sufficient bandwidth, io bound things like database access and JSP compiles will be faster.

I'd imagine that your (new?) dual athlon has lots of fast disk on 160MB SCSI channels. That's just gonna rip (relatively) no matter what the software is doing.

I'm not disagreeing that there may be problems with certain kernel builds and java VMs threading on linux SMP - I'm just urging you not to jump to conclusions when the hardware is so different.

Gary S. Cuozzo wrote:

Funny you should mention that. My usual desktop is a dual P3-700 SCSI box, but has a very old (read slow) IBM disk in it. I was working last night on my laptop (IBM Thinkpad, ~800MHz) and refreshed my schema and thought "wow, that was WAY faster than my desktop". JSP pages & such seem to compile much faster too. I'm running the exact same s/w setup as you are on both boxes.

However, we just went to production on a dual AMD Athlon MP, running debian not RH, and that box smokes. We have not had a big load on it yet, but the performance difference is noticable.

This leads me to believe that maybe there are tweaks that could/should be made to the stock RH kernel. Right now, I'm running 2.4.18-3smp installed via RPM's. I'm considering trying the enterprise kernel to see if there is a difference. Beyond that, I guess I could try custom building one and try that too. Like I said, I have a very old & slow disk in that system, but even in-memory apps don't seem to run as well as my laptop.

Please let us know if you find anything concrete, I'll do the same.
gary.

Darren Hartford wrote:

Hey all,
I have been having some performance issues with JBoss, and thought it related to ignorance or poor CMP/cache setup. However, I just happen to test the same application from a dual-processor Linux box to a single-processor Linux laptop, and the laptop performed better. Also have a specialized java-based XML-RPC application server (not related to JBoss) that has the same issue of performance between dual-processor and single-processor hardware servers.

All are running Redhat 7.3, running Sun JDK 1.4.1_01. Any ideas if java on Linux SMP has some performance issues, or if someone has had similar issues?

-D



-------------------------------------------------------
This SF.NET email is sponsored by: FREE  SSL Guide from Thawte
are you planning your Web Server Security? Click here to get a FREE
Thawte SSL guide and find the answers to all your  SSL security issues.
http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0026en
_______________________________________________
JBoss-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-user




-------------------------------------------------------
This SF.NET email is sponsored by: Take your first step towards giving your online business a competitive advantage. Test-drive a Thawte SSL certificate - our easy online guide will show you how. Click here to get started: http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0027en
_______________________________________________
JBoss-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-user

Reply via email to