The invention is suitable for the technical field of large-scale data processing and particularly relates to a data aggregation query method and apparatus. The method comprises: when receiving a query request for a database cluster, determining a hash table corresponding to the query request, determining a plurality of partition tables corresponding to the hash table, and generating a MapReduce query task; through scheduling nodes in an Hadoop Yarn framework, according to the MapReduce query task, determining a plurality of subtasks, and distributing the subtasks to a plurality of computing nodes; through the computing nodes, performing the subtasks, obtaining a plurality of computing results, and through the computing nodes, feeding the computing results back to the scheduling nodes; and through the scheduling nodes, simplifying the computing results, and obtaining a query result corresponding to the query request. The method and the apparatus realize relational query and statistics of the relevant partition tables in the database cluster, reduce network overhead, and improve data aggregation query efficiency.