I'm a big fan of doing comparative performance tests on [u]identical hardware[/u]. This eliminates a whole bunch of pesky extraneous variables, which can mess up (bias) a performance comparison.
My pet peeve: I've seen ads in the past (I won't say which company created them!) that said "Solaris is faster than Linux". Then in the fine print you read that Linux was running on a Xeon box, and Solaris was running on an Opteron box (with different amounts of memory, different hard drives, etc.). It may be true that one is faster than the other, but *that* kind of test doesn't prove it either way. Perhaps the ad should have said something like "Behold our box running Solaris that runs faster than the competition's box running Linux!" Call me pedantic, but I think this statement is more accurate. OK, so I'll get off my soapbox now... :-) Here's an interesting comparison, done by the textdrive.com folks. I thought that the OpenSolaris performance community might be interested. http://weblog.textdrive.com/article/220/comparative-netperf-network-performance-of-freebsd-versus-solaris-on-identical-hardware So, what do other people use for benchmarks in this area, besides netperf? This message posted from opensolaris.org _______________________________________________ perf-discuss mailing list perf-discuss@opensolaris.org
