I checked the website (rubini.us) for some guidance on how to profile ruby code under rbx, but I came up empty.
I try to run most of my day-to-day development code under rbx (works great most of the time). Lately I've been doing quite a bit of work with mongodb. I went so far as to contribute patches to both rbx and mongodb to get the C extension compiling and working under both. However, all of my ruby code that interfaces to mongo is *really* slow even with the C extension. To give myself some idea of the difference, I have run benchmarks comparing ruby 1.9.1 to rbx using the official mongo ruby driver as well as em-mongo (evented driver). Ruby 1.9.1 can insert at a rate of over 1000 docs per second whereas rbx inserts around 20/sec. (Here's a copy of my little benchmark program.) http://gist.github.com/371389 I have some ideas about where the bottlenecks are but no real proof. I can easily profile this code under ruby 1.8.x or 1.9.x but I don't know how to do that for rbx. I would be grateful for any pointers. Thanks... cr -- --- !ruby/object:MailingList name: rubinius-dev view: http://groups.google.com/group/rubinius-dev?hl=en post: [email protected] unsubscribe: [email protected]
