I have recently added a number of QA tests to our Hudson CI build job. The number of QA tests run has increased from 40-something (the joinmanager category) to 337 (id, loader, policyprovider, locatordiscovery, activation, config, discoverymanager, joinmanager, url, iiop, jrmp, reliability, thread categories)
Needless to say the build takes a lot longer (it takes hours now)... Would it make sense to split up into two separate builds? To make sure - one build job that polls SVN hourly, builds, runs unit (i.e. not QA nor jtreg) tests and packages up. We could call this the "release" build: River-release. - one build job that is scheduled daily, builds, runs unit tests, and runs the QA (and soon jtreg) tests (no packaging here). We could call this the "qa" build: River-qa. Thoughts? Thanks Jonathan
