Batch insertion strategy in a distribution database

History data produced by financial enterprises are usually very large, and needed to be transmitted from one table partly or wholly to another table, which requests flexible and efficient insertion strategy. Distributed systems are good at handling massive data under big data era. For the high performance and extendibility, financial enterprises recently tend to handle and analyze their business data with distributed systems instead of traditional ‘IOE’ architecture. Distributed systems are weak at SQL support, like HBase, which only provides some simple program interfaces for users, not satisfying financial enterprise's data insertion requirement. To solve these problems, we proposes a batch insertion strategy(BIS). The main contents of BIS include multiple insertion strategies used to implement large data inserting, threshold optimization technology used to decrease the network cost and redirection technology used to reduce the pressure of system. BIS is implemented in a distributed database system called OceanBase which is designed by Alibaba Group. The experiment data are from the actual business data of some financial enterprises, and the experiment results show that performance of BIS is basically as well as existed value insertion in OceanBase but much better than program insertion.