Computing power resource allocation method and device, equipment and storage medium

Abstract

The embodiment of the invention discloses a computing power resource allocation method and device, equipment and a storage medium, containers corresponding to different business configurations, a computing power resource quota corresponding to each container configuration converts a first instance into an instance task when the submitted first instance is obtained for a target business, and according to a corresponding relationship between the business and the container, the target container corresponding to the target computing is determined. Iif it is determined that no available computing power resource exists in the target container according to the computing power resource quota corresponding to the target container, the instance task corresponding to the first instance is controlledto enter a task queue of the target container. When it is monitored that the available computing power resources appear in the target container, the target instance task is scheduled from the task queue to be put into operation. Since different services are isolated through the container, examples of the same service share the computing power resources of the same container, even if priority preemption occurs during scheduling, the priority preemption is only constrained in the container, the problem of computing power resource preemption among different services is avoided, and the user experience is improved.

Description

technical field [0001] The present application relates to the field of data processing, and in particular to a computing power resource allocation method, device, equipment and storage medium. Background technique [0002] With the rapid development of science and technology, various advanced technologies continue to emerge. A Graphics Processing Unit (GPU) is becoming more and more popular due to its good computing capability. GPUs are often used for computational processing in various scenarios. For example, it is used for artificial intelligence (AI) model training. [0003] During the training process of AI models for different businesses, computing resources need to be shared. For the use of computing power resources for business, the current solution is to provide a unified task layer, and multiple business instances are placed in the unified task layer queue to wait for scheduling. High and low priorities preempt computing power resources. [0004] However, this ...

AI Technical Summary

Problems solved by technology

[0004] However, this method expands the scope of priority preemption scheduling among different businesses, which leads to the problem of preemption of computing resources between different businesses and affects user experience.

Images