Hi,
I am very happy that I managed to create a very small deploy image based on
Pharo-1.1-11400-rc2. This image contains all of Seaside and Glorp/PostgreSQL
and is still only about 12Mb. All this thanks to ScriptLoader's
#cleanUpForProduction.
Thanks a lot to all who made this possible.
I noticed however that this image is handling my HTTP/Seaside benchmark
consistenly faster on my Mac, about 80 to 85 req/s (Mac OS X 10.6.4, 2.4 Ghz
Intel Core 2 Duo with 4GB RAM) than on my server (RHEL 4 (update 7), 2 Ghz AMD
Opteron 175 Dual Core with 4GB RAM), about 55 to 65 req/s.
On my Mac I use the Squeak 4.2.4beta1U VM, on my server the Squeak-4.0.3.2202
Unix VM.
Next I tried the Squeak-4.0.3.2202 VM on my Mac. This resulted in about 60 to
70 req/s.
Is it normal that the Unix VM is slower than the Mac VM ?
Shouldn't a Linux server box be faster at networking/IO/CPU than a MacBook Pro
running the full Mac OS X GUI ?
Is there anything simple (low hanging fruit kind of stuff) that I can do to
improve performance, either using command line options to the VM or using
settings at the image level ?
I currently run the VM with:
Squeak-4.0.3.2202-linux_i386/bin/squeak -vm-display-null -vm-sound-null
deploy.image
What are others using in the Seaside deploy area ?
Thanks,
Sven
_______________________________________________
Pharo-users mailing list
[email protected]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-users
