Search

What is the need for serialization in Mapreduce ?



Below are the two necessities for serialization in Hadoop.
  • In Hadoop cluster, data is stored in only binary stream format but object structured data can’t be stored directly hadoop data nodes.
  • Only Binary stream data can be transferred across data nodes in hadoop cluster. So, Serialization is needed to convert the object structured data into binary stream format.
Posted by at
Labels:
  • What are the main components of Mapreduce Job ?
      Main driver class which provides job configuration parameters. Mapper class which must extend org.apache.hadoop.mapredu...
  • TO_CHAR
       This will be used to extract various date formats.     The available date formats as follows.     Syntax: to_char ( date , fo...