Existing distributed database multi-connection optimization searches have high space complexity, which greatly affects the search efficiency. In this paper, we propose a parallel genetic-maximum minimum ant colony algorithm of (PGA-MMAS) for query optimization. Firstly, based on the faster convergence of genetic algorithm (GA), initialization coding and optimal population screening operations are performed on the final connectivity relations to obtain a set of better query execution plan (QEP), and the better QEP is transformed into the initialized pheromone distribution of maximal minimal ant colony algorithm, and the updating and cycling of the pheromone matrix is performed according to pheromone updating rules, and the global optimal QEP is finally searched more rapidly. The improved crossover operation of this paper’s optimization algorithm improves the solution efficiency by 10%, and the total execution time is also greatly reduced compared with the three compared algorithms. The query latency of this paper’s algorithm is reduced compared to the SparkSQL algorithm on five different experiments, and the reduction is 60% and above. It shows that the algorithm in this paper can improve the query efficiency of distributed databases and ensure a more efficient query method.
