tag:blogger.com,1999:blog-6406593750327945950.post4814098732920095589..comments2022-12-04T20:47:13.182-08:00Comments on Igor Minar's Blog: Benchmarking JRuby on RailsIgor Minarhttp://www.blogger.com/profile/03520548417275543432noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-6406593750327945950.post-67789155394185500942009-02-17T07:12:00.000-08:002009-02-17T07:12:00.000-08:00Was the number of requests per second the same in ...Was the number of requests per second the same in both cases? The difference in cpu usage only shows that jruby can use all four cpu cores you have, while c-ruby can use only one.Igor Minarhttps://www.blogger.com/profile/03520548417275543432noreply@blogger.comtag:blogger.com,1999:blog-6406593750327945950.post-29533243712178415052009-02-17T01:34:00.000-08:002009-02-17T01:34:00.000-08:00I benchmarked jruby-1.1.6, glassfish3 and rails-2....I benchmarked jruby-1.1.6, glassfish3 and rails-2.2 running on threadsafe mode.<BR/><BR/>The page I tested had several db lookups (perhaps 20), so it was rather rough on the implementation. I noticed that jruby+glassfish with several threads (i.e. stressing the CPU at 400%) was as fast as ruby+mongrel in single thread (i.e. stressing the CPU at 100%). So, jruby+glassfish was 4 times slower than ruby+mongrel. The DB used was mysql, and on jruby, I used the jdbcmysql adapter.Anonymousnoreply@blogger.com