I fully believe that complicated tests *need* to be run to run to find OS issues. There are plenty of applications that people are running and benchmarks and workloads that will find these, however many of these are not open ex: SPEC, TPC, SAP, Oracle Apps, Manugistics, SAS, Informatica, Oracle, DB2, Sybase,.... OK so various partners of these organizations will have to work directly with these groups on performance and then fix them in Open Solaris.
I worry that just searching the web for open applications that seem to be complex as real ones may not be as realistic as real applications. Again I say it would be best to focus on making lower level tests really good, as opposed to trying to find, create, modify, administer "Open" complex application benchmark test suites. I would hate everyone to spend lots of time creating poor-man's SPEC suite (or whatever). High-Quality tests at each level should be the goal to producing the best OpenSolaris. If some benchmarks are not open and they provide the best tests than the various developers who have relationships and access to these applications should work them, I think this would be better than just using any application found on the web. This message posted from opensolaris.org _______________________________________________ perf-discuss mailing list perf-discuss@opensolaris.org
