You are right that the -T option and surefire's parallelism are complementary. Given equal-sized modules the effect tweaking surefire vs -T parallelism should be similar.
The fork/thread related parameters to surefire have NO relation to the -T parameters, which means "reuseForks" is ineffective in your settings. I suspect you're simply trashing your os or the VM. Look at disk/memory usage indicators OS level. In terms of resource efficiency there's a clear difference between the different alternatives: A) Multiple threads within same (or a single forked) VM inside surefire. By far the most resource efficient (Probably 10x faster than the alternatives) B) Multiple surefire forks (Can be resource heavy too !) C) -T n option, which can amplify other settings "n-fold". Kristian 2014-10-08 21:11 GMT+02:00 laredotornado-3 <[email protected]>: > Hi, > > Karl, to answer your question, I only have one processor on my MacBook Pro > (10.9.5), which is a 2.2 GHz Intel Core i7. I haven't set "forkCount" in > either my surefire or failsafe plugin (configs listed in question) because I > was under the impression that the "-T" option took care of that. > > Jörg, My $MAVEN_OPTS var is set to -Xmx4096m, but I thought I read somewhere > that in the case of surefire/failsafe, MAVEN_OPTS is ignored in favor of the > configurations listed in the plugins, which are "-Xmx2048m > -XX:MaxPermSize=512M" for surefire and "-Xmx4096m -XX:MaxPermSize=512M" for > failsafe. > > Thanks, - Dave > > > > -- > View this message in context: > http://maven.40175.n5.nabble.com/Why-do-running-modules-in-parallel-take-longer-to-execute-than-running-them-in-sequence-tp5808128p5808146.html > Sent from the Maven - Users mailing list archive at Nabble.com. > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
