Unlock instant, AI-driven research and patent intelligence for your innovation.

Resource load balancing control method and cluster scheduler

A load balancing and control method technology, applied in the field of communication, can solve problems such as wrong decision-making, incorrect prediction of load data, and inability to control the load balancing of cluster system resources.

Active Publication Date: 2021-10-22
HUAWEI TECH CO LTD
View PDF7 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, among the above two methods, the first method will produce invalid migration in some cases, for example: the load of the cluster nodes is high at T1, and the load of the cluster nodes returns to a reasonable level at T1+1. In this case, no virtual machine migration is required
In the second method, the cluster node is used as the carrier of the virtual machine, and the virtual machine deployed on the cluster node changes, so the load data of the cluster node does not have regularity
Therefore, making predictions based on the load data of the cluster nodes is incorrect and may lead to wrong decisions
[0004] It can be seen that neither of the above two methods can accurately balance the resource load of the cluster system.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Resource load balancing control method and cluster scheduler
  • Resource load balancing control method and cluster scheduler
  • Resource load balancing control method and cluster scheduler

Examples

Experimental program
Comparison scheme
Effect test

Embodiment approach

[0073] As another optional implementation manner, the method also includes the following steps:

[0074] 21) For each cluster node, construct training data using historical performance data of applications deployed on the cluster node;

[0075] 22) Use the training data for training to obtain a data model.

[0076] Specifically, the manner in which the cluster scheduler predicts the performance data of the application deployed on each cluster node within a preset time period may specifically be:

[0077] 31) For each cluster node, determine multiple applications deployed on the cluster node;

[0078] 32) For each application, construct historical performance data of the application based on the current cycle as input data;

[0079] 33) Substituting the input data into the data model to predict the performance data of the application within a preset time period;

[0080] 34) Add the performance data of all applications predicted on the cluster nodes to obtain the performance...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The embodiment of the invention discloses a resource load balancing control method and a cluster scheduler. Wherein, the method includes: predicting the performance data of the application deployed on each of the cluster nodes within a preset time period; calculating the first standard deviation of the cluster system according to the predicted performance data of each of the cluster nodes ; When the first standard deviation of the cluster system is greater than a preset threshold, determine an application migration plan according to resource load balancing rules; send the application migration plan to the cluster application manager to trigger the cluster The application manager balances and controls the resource load of the cluster system according to the application migration scheme. The implementation of the embodiments of the present invention can improve the effectiveness of the migration, and at the same time, more accurately balance and control the resource load of the cluster system.

Description

technical field [0001] The invention relates to the field of communication technology, in particular to a resource load balancing control method and a cluster scheduler. Background technique [0002] With the development of virtualization and cloud computing, cloud centers are increasingly used. The cloud center includes at least one cluster system, and each cluster system includes a cluster scheduler, a cluster application manager, and multiple cluster nodes. After the cluster scheduler receives the application deployment request, it can deploy the application to different cluster nodes. In order to make the use of resources (such as computing resources, storage resources, and network resources) more balanced and reasonable, the cluster application manager needs to monitor the resource load of each cluster node to balance the resource load of the cluster system. [0003] Currently, there are two ways to balance and control the resource load of the cluster system. The fir...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06F9/50
CPCG06F9/5088G06F2209/5022H04L67/1008H04L67/1029
Inventor 杜树宝金锡波
Owner HUAWEI TECH CO LTD