Method for parallel query processing of data comprising a replica in distributed database

摘要

The present invention relates to a technique for reducing a query performance time for large-capacity distributed data having replicas existing by changing a query for data having replicas existing in a distributed database to a plurality of range condition queries in correspondence to the number of nodes in which the replicas are stored, and simultaneously performing the range condition queries in a plurality of nodes. A method for paralleling queries for a table having replicas existing in a distributed database comprises: a first step of allowing a master server to determine a query as a target query to be divided query when a target table includes a column capable of designating a range while being a distributed table by analyzing the query requested from a client in a master server; a second step of allowing the master server to determine whether the number of target records to be searched for the target query to be divided exceeds the preset number of reference records; a third step of allowing the master server to check slave servers having an original and a copy when the number of the target records to be searched exceeds the preset number of reference records in the second step; a fourth step of allowing the master server to generate a range query on each slave server by dividing a record area based on the number of target records to be searched and the number of slave servers having the copy; and a fifth step of allowing the master server to generate a result for the query requested from the client by simultaneously transmitting the range query to each slave server in parallel and collecting execution result data of the range query from each slave server to combine the collected execution result data.