22 / 58

We have a cluster having two racks R1 and R2. R1 has three machines M1, M2 and M3. R2 has three machines M4, M5 and M6. On M1 we have namenode only. We have installed data node on rest of the machines. If we have logged in M3 and we have a file of size 100MB on the local file system of M3. The default replication factor is 3 and all datanodes have ample amount of space available. If we copy this file to HDFS having 3 replication factor, which nodes are most likely to have the replicas of this files' data?