How is an HDFS chunk different from a Spark RDD$?$ An HDFS chunk resides on a hard disk while a Spark RDD typically resides in memory $($RAM$)$ An HDFS chunk is split between several nodes while a Spark RDD is stored only on one node A Spark RDD is fault tolerant while an HDFS chunk is not An HDFS chunk resides on the master node while a Spark RDD resides on a compute node