This chapter describes how to allocate shared CPU and memory resources among users and groups in a Hadoop cluster.

CPU scheduling represents just one aspect of YARN resource management, which also includes CGroups, node labels, archival storage, and memory as storage. CGroups should be used with CPU scheduling to constrain and manage CPU processes.

	Note
	You should use CPU scheduling only in a Linux environment, because there is no isolation mechanism (CGroups equivalent) for Windows.