Batch Query Partitioning for Multi-engine Execution

원문 (중국어)

Techniques are disclosed relating to partitioning batch queries for multi-engine execution. A system receives a batch query specifying query data and including a request for geospatial data for regions corresponding to the query data. Based on locations corresponding to the query data, the system partitions the query data into subsets. The system may assign the subsets of query data to query engines corresponding to the locations of the subsets. The system may cause the engines to retrieve geographic region data corresponding to the locations included in the subsets, where the retrieving is performed by a given query engine for a corresponding subset by accessing an in-memory index of the given engine that stores geographic region data for a geographic partition within which the corresponding subset of query data is located. The system may store region data retrieved by the engines for the subsets in an aggregated data store.