Hi all, It is know that the manual said the number of parallel tasks should be set corresponding to the number of cores. If one task is scheduled to one core, how it is bound by TaskTracker?
Are there any work having been done in Hadoop, to try to schedule task based on core? Thanks, Ian
