Cae Fernandes wrote: > Hello guys, > > I would like to know if anybody could provide me with a unbiased "tomcat vs > others" peformance benchmark with full tests and full system specs > (hardware, os, etc). > > I'm interested in concurrent amount of users, hardware load and specs. > I've been designing a new system that may need to sustaint a heavy amount of > users and I need to know what to expect. > > Any help is apreciated.
Performance will be 98% your app, 2% the container. If you get to the point where the container is the major differentiating factor in performance then either a) you have the container configured badly or b) you are close to getting as much performance as you can out of one instance of your app and you'd be better off looking at options for scaling. Some related observations: - Benchmarks are pretty meaningless. You can construct a benchmark to show pretty much whatever you want. For example, I've seen benchmarks that 'prove' that each of the BIO, NIO and APR/native connector gives the 'best' performance. - The only benchmark that matters is how your app performs - A good profiler is your friend. I use YourKit (because they give open source committers free copies) Mark --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
