Scheduling method, device, equipment, storage medium and product for server cluster

By receiving and counting service requests, underutilized nodes were identified. Service type and Q-learning network strategies were used to optimize node allocation in the server cluster, solving the problem of resource waste and improving the service performance of the server cluster.

CN119922192BActive Publication Date: 2025-11-04CHINA MOBILE (SUZHOU) SOFTWARE TECH CO LTD +1
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202510073134.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-01-16
Publication Date
2025-11-04
Estimated Expiration
2045-01-16

AI Technical Summary

Technical Problem

Traditional dynamic scheduling algorithms can easily lead to resource waste in server clusters because different users use different types of services and have different resource consumption, resulting in inappropriate allocation.

Method used

By receiving service requests and counting the number of service requests within a set time period, the latest underutilized nodes in the server cluster are identified. Based on the comparison results of service type and number of nodes, a target scheduling strategy is adopted to allocate service nodes to service requests, including a first target scheduling strategy based on service type and a second target scheduling strategy based on Q-learning network. The node allocation is optimized by combining the reinforcement learning algorithm of Q-learning network.

Benefits of technology

This approach optimizes service node allocation, reduces resource waste, and improves the service performance of server clusters by taking into account the differences in resource consumption among different service types.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119922192B_ABST
    Figure CN119922192B_ABST
Patent Text Reader

Abstract

The application discloses a scheduling method and device for a server cluster, equipment, a storage medium and a product. The method comprises the following steps: receiving a service request and counting a first number of service requests within a set time period; each service request comprises at least first information indicating a service type; determining a second number of the latest underutilized nodes in the server cluster; determining a target scheduling strategy of the server cluster based on a comparison result of the first number and the second number; the target scheduling strategy comprises evaluation parameters corresponding to the service type; and allocating service nodes for each service request based on the target scheduling strategy. The service nodes can be allocated for the service request in the server cluster based on comprehensive consideration of the service type and the comparison result of the first number and the second number, so that the optimized allocation of the service nodes can be realized based on consideration of the resource consumption difference of different service types, and then the resource waste can be reduced and the service performance of the server cluster can be improved.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Kubernetes resource scheduling method and device and electronic equipment

    CN116991589A

  • TR-DQN-based high-performance computing cluster resource scheduling method and system

    CN117591273A