Optimization method, device, terminal equipment and storage medium for multi-tenant resource scheduling

Abstract

The invention discloses an optimization method, device, terminal equipment and storage medium for multi-tenant resource scheduling. The optimization method for multi-tenant resource scheduling includes: receiving a resource scheduling request required by each tenant's task to be executed; The size of the required resources is divided into large tasks and small tasks, and the proportion of the concurrently running large tasks and small tasks in the pending tasks of the tenant is adjusted according to the total amount of resources allocated to each tenant; according to the adjusted simultaneous running According to the proportion of large tasks and small tasks in each tenant, the virtual resource allocation is performed on the large tasks and small tasks of each tenant. The optimization method for multi-tenant resource scheduling of the present invention improves the utilization rate of system resources.

Description

Technical field[0001]The present invention relates to a resource management technique, and more particularly to a multi-tenant resource scheduling optimization method, a device, a terminal device, and a computer readable storage medium.Background technique[0002]With the arrival of Internet big data age, distributed system Hadoop plays an important role in information engineering such as industry construction data, large data platforms. With the data center, the data accumulation of large data platforms, the promotion of capacity-opening work, will face multiple organizations or institutions while applying production on the platform. In order to solve the manager's use management of each user (Central Processing Unit, CPU), memory, etc.), the industry puts forward multi-tenant concept, that is, managers can put resources in distributed systems according to each The actual usage of users is allocated and adjusted, and the platform technology needs to provide resources between users, a...

AI Technical Summary

Problems solved by technology

For the case where the decomposed small data blocks run more, because the start time of each small data block is different or the resources consumed in the process of mapping (map) and reduction (reduce) are quite different, if the resources used simply by task If the amount of task scheduling is performed, some resources will usually be idle at the end of the task operation, resulting in a decrease in the utilization of system resources.

Images