The users can be created using below steps: a)Get the information from user as to which machine is he working from. b)create the user in in OS first. c)Create the user in Hadoop by creating his home folder /user/username in Hadoop d)make sure that we have 777 permission for temp directory in HDFS e)using chown command change ownership from Hadoop to user for only his home directory so that he can write into only his directory and not other users. f)add his name into name node hdfs dfsadmin -refreshUserToGroupMappings G)If needed set a space limit for the user to limit the amount of data stored by him.hdfs dfsadmin -setSpaceQuota 50g /user/username ... View more

@rinu shrivastav The split size is calculated by the formula:- max(mapred.min.split.size, min(mapred.max.split.size, dfs.block.size)) Say, HDFS block size is 64 MB and min.input.size is set to 128MB, then there will be split size would be 128MB. To read 256MB of data, there will be two mappers. To increase the number of mappers, then you could decrease min.input.size till the HDFS block size. split size=max(128,min(256,64)) ... View more

TaskTracker & JobTracker doesn't exist with YARN. The default replication factor is 3. ... View more