Task execution method and system based on distributed system

Abstract

The invention discloses a task execution method based on a distributed system. The distributed system comprises a management end and a plurality of execution ends, wherein the management end is used for sending a received task to a selected first execution end, wherein the task comprises a plurality of task modules allowing to be executed in sequence and at least comprises a packaging module packaging the task module capable of being sliced; when the task is executed at the first execution end, the method comprises the following steps of unsealing the packaging module and sending a slicing request and original information of the unsealed task module to the management end; the management end is used for slicing the task module into a plurality of slice tasks according to the original information of the task module and the quantity of available second execution ends, and sending the plurality of slice tasks to the first execution end and the plurality of second execution ends; the corresponding slice tasks are executed by the first execution end and the plurality of second execution ends. The task execution method based on the distributed system has a clear flow; the execution time of the tasks is shortened and existing resources are sufficiently utilized.

Description

technical field [0001] The invention relates to the technical field of radio and television broadcasting systems, in particular to a task execution method and system based on a distributed system. Background technique [0002] At present, most of the equipment systems of radio and television stations are distributed computer systems, which usually perform corresponding tasks to complete a certain job. The content of work requirements is different, the tasks are different, and the resources they occupy are also different. For example, material migration tasks mainly occupy disk I / O and network I / O, and transcoding and file technical review tasks mainly occupy CPU, memory, disk I / O, and network I / O resources. When a certain computer (hereinafter referred to as the execution end) in the system is performing a task, if the task takes a long time, it will be in the working state for a long time, while the other execution ends are idle state. This creates two problems: [0003...

AI Technical Summary

Problems solved by technology

In an emergency state, it is hoped that the technical review can be completed in the shortest time, but this single-machine execution method cannot meet the requirements

[0004] Second, resources are not fully utilized, which indirectly causes waste of resources

Due to the different lengths of the tasks, when the tasks are allocated, multiple execution ends will be idle, and one execution end will be in a long-term working state, so the existing resources cannot be fully utilized.

Images