On Mon, June 19, 2006 10:08 am, Dimitri Kovalov wrote: > I have a test application that loads a database with about > 20K records. Just sequential writes and I get very strange > results. With 1 CPU it take :49 and with 2 cpu it take > 2:55. Is this possible? Its a slow disk but I surprized by > results.
MySQL performance is historically poorer under FreeBSD, and we probably inherited the general malaise. (Note I'm saying that without benchmarking it myself...) You may want to look at: http://software.newsforge.com/article.pl?sid=04/12/27/1243207 http://www.mysqlperformanceblog.com/2006/06/15/freebsd-tests/ http://wikitest.freebsd.org/MySQL Interestingly, that second link shows much better results with a 4BSD-style scheduler, which is what we pretty much have now.
