Queries per second (QPS) or transactions per second (TPS) and response time (RT) are essential metrics for measuring the performance of a database system. QPS/TPS indicates the number of queries or transactions that the system can process per second, and RT is the time the system takes to process a request. The resources of a single server can be fully utilized by improving the QPS/TPS, thereby improving the cost effectiveness and reducing the total costs of ownership (TCOs) of database servers. The QPS and response speed of upstream applications can be improved by reducing the RT, thereby optimizing user experience and further improving the QPS/TPS of the database system.
Performance tuning is about using technical methods to improve the QPS/TPS and reduce the RT. Generally, a performance tuning cycle is completed in the following three steps: determine the optimization direction > find the performance bottleneck > prepare and execute the optimization solution. Performance tuning is a systematic work that involves technology stacks such as hardware, operating systems, and software. This requires you to master its core principles and tuning tools.
OceanBase Database is a distributed database management system. In the previous chapters, we have introduced the high scalability and high availability features of the system. In this chapter about high performance, we will introduce you to the techniques for performance tuning of OceanBase Database.