Dispatching method and system for multi-priority rendering resources based on message queues

Abstract

The invention relates to a dispatching method and system for multi-priority rendering resources based on message queues. The method comprises the steps of A, grouping task queues according to different types; B, storing task data in a database and generating task IDs, wherein the task IDs are progressively increased from small to large according to the sequence of the tasks; C, actively connecting with each task queue by a rendering agent, and dragging rendering task data from the task queues for rendering; D, judging whether there are tasks queuing in the task queues; E, dragging the tasks from the corresponding task queues for rendering by the rendering agent according to configuration after the rendering is finished, and performing the step D; F, searching other task queues and judging whether there are queuing tasks; and G, dragging the tasks from the corresponding task queues for rendering by the rendering agent according to configuration, and performing the step D. According to the multi-priority queue dispatching strategy, the emergency tasks can be handled preferentially, the rendering resources of the secondary tasks will not be deprived continuously, and the strategy is more flexible and smooth.

Description

technical field [0001] The invention belongs to the field of improving message concurrency technology, and in particular relates to a scheduling method and system for multi-priority rendering resources based on message queues. Background technique [0002] In order to improve the rendering speed and the amount of concurrent rendering, no matter how high the performance of a single rendering server is, there is a bottleneck: if it supports multiple concurrent rendering, it will inevitably increase the time-consuming of each rendering task. At this time, a rendering cluster composed of dozens or even hundreds of rendering servers can increase the amount of concurrent rendering while maintaining a high rendering speed, and support multiple users to render at the same time. [0003] In order to maximize the resources of each rendering server and meet the changing needs of users, an efficient and flexible rendering resource scheduling scheme is very important. [0004] The gener...

AI Technical Summary

Problems solved by technology

[0005] Single task queue, unable to flexibly handle multi-priority rendering tasks

Priority tasks are often inserted directly into the head of the queue, which is relatively simple and rude. If there are many priority tasks, the secondary tasks will never get rendering resources

[0006] Two long-time-consuming rendering tasks are rendered on the same machine at the same time, competing for computing resources, resulting in two long-time-consuming rendering tasks taking longer to complete rendering

Increase and offline and restart render server complexity

Images