The invention discloses an automatic optimization method of a relational database on the basis of a cloud platform, and belongs to the technical field of cloud platform database optimization. The method can collect the operation state information of a database and the hardware index information of a virtual machine where the database is installed in fixed time, can timely output an initiative optimization suggestion, a passive optimization suggestion and automatic passive optimization, and improves the operation performance stability of the database in the cloud platform. According to a database coding protocol, a workprocess process plugin code required by a workprocess process is edited, an item directory is created under a contrib directory of a database source code directory, and the workprocess process plugin code is copied and installed into the item directory; a workprocess process plugin is created in the database; a database configuration file postgresql.conf is configured, and a dynamic-link library file corresponding to the workprocess process is loaded; and when the database starts, a _PG_init hook program and a workprocess process entry function are called to start the workprocess process. The automatic optimization method can output the initiative optimization suggestion, the passive optimization suggestion and the automatic passive optimization.