Skip to content
Miguel Duarte edited this page Apr 5, 2015 · 5 revisions

The --taskexecutor argument indicates how the experiments will run. Three options are available: SequentialTaskExecutor, ParallelTaskExecutor, and ConillonTaskExecutor. Specify the task executor you want to run like so:

--taskexecutor classname=ParallelTaskExecutor

  • SequentialTaskExecutor: Each sample will run sequentially, one after the other, using just one CPU core.
  • ParallelTaskExecutor: All CPU cores will be used by executing several samples in parallel.
  • ConillonTaskExecutor: All tasks will be executed on Conillon instead of the host computer. Check the Conillon page for more information. It is necessary to specify the server address (--taskexecutor classname=ConillonTaskExecutor,server=evolve.dcti.iscte.pt)