We explored parallel performance in solaris and sun servers, and among others did some benchmarks on real life tasks by running pbzip with a large file (snip from middle of an iso image) to tmpfs or devnull directly.
one trick is to monitor output file creation rate with a single threaded run as to pick a dataset with linear compression rate. got very interesting results on niagara cores vs strands this way. quite cool. for web you can try olio and faban jim from mobile PS Ben, you're such a guru, do you also need to ask? ,) -- This message posted from opensolaris.org _______________________________________________ sysadmin-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/sysadmin-discuss
