基于多线程并行强化学习的数据库索引推荐

Database index recommendation based on multi-thread parallelreinforcement learning

建立索引是提高数据库性能的一个重要方法。目前随着强化学习算法的发展,出现了一系列使用强化学习解决索引推荐问题(index selection problem, ISP)的方法。针对现有的深度强化学习索引推荐算法训练时间长、训练不够稳定的问题,提出了一个基于A2C的索引推荐算法PRELIA。该算法加入负载索引扫描行数特征矩阵,并对奖励值进行归一化处理,旨在提高索引选择的准确性和效率,减少索引空间占用。在不同数据集上的实验结果表示,该算法可以在保证与比较算法相当的索引推荐质量的同时,推荐出的索引占用更小的存储空间,其训练时间比基线算法时间提高了4倍以上。

Indexing is an important method to improve database performance.At present,with the development of reinforcement learning algorithm,there are a series of methods to solve the index recommendation problem by reinforcement learning.Aiming at the problem that the existing deep reinforcement learning index recommendation algorithm has long training time and unstable training,this paper proposed an index recommendation algorithm based on A2C(advantage actor-critical),called PRELIA(parallel compensation learning index advisor).In order to improve the accuracy and efficiency of index selection and reduce the occupation of index space,the algorithm added the characteristic matrix of the number of rows scanned by load index and normalized the reward value.Experimental results on different data sets show that the proposed algorithm can gua-rantee the index recommendation quality equivalent to that of the compared algorithms,while the recommended index occupies less storage space,and the training time is more than 4 times longer than that of the baseline algorithms.