Task scheduling method and application based on cloud edge fusion server network architecture

The invention discloses a task scheduling method and application based on a cloud edge fusion server network architecture. The method comprises the steps that a target slave server queries whether complete data corresponding to a current task allocated by a master server exists locally or not; if yes, the target slave server arranges the current task into a waiting queue of the target slave serverto wait for execution; and if not, the target slave server arranges the current task into the preset queue of the target slave server, and waits for the corresponding data to complete transmission. According to the cloud edge fusion server network architecture, a task scheduling mechanism is configured based on a waiting queue and a pre-prepared queue in an edge computing layer. Distributed parallel computing can be carried out by simultaneously combining a plurality of edge servers distributed at different geographic positions, so that the aim of processing low-delay response when jobs submitted by a plurality of users are processed on an edge computing layer is fulfilled, and the data processing efficiency of a server network is optimized.