MapReduce in Hadoop maintains some built-in counters for every job. These counters report various metrics. There are counters for the number of bytes and records that allows us to confirm that the expected amount of input is consumed and the expected amount of output is produced.
These built-in counters are: