RECORD PROCESSING METHOD USING INDEX DATA STRUCTURE IN DISTRIBUTED PROCESSING SYSTEM BASED ON MAPREDUCE

摘要

A method for processing a record by using an index in a MapReduce-based distribution processing system comprises: a step in which a distribution node of a distribution processing system classifies records as analysis targets in an input file; a step in which the distribution node generates a data structure having a plurality of indexes indicating a key and a storage location of each record; a step in which the distribution node generates a new index data structure having a new index indicating a key of a record, a storage location of a record, and an identifier of a data structure to which a record belongs while approaching the plurality of indexes in order of the keys in the data structure; and a step in which the distribution node applies a reduce function to a record stored on the basis of an identifier of a data structure and a storage location of a record indicated by an index having the same key while approaching indexes in order of keys in the new data structure.