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

Cluster load prediction method and distributed cluster management system

A distributed cluster and load prediction technology, applied in transmission systems, electrical components, etc., can solve problems such as lack of flexibility and inability to adaptively select algorithms, and achieve the effect of strong scalability and strong algorithm controllability

Active Publication Date: 2014-06-18
ZHEJIANG UNIV
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] The present invention aims to solve the defects of the above-mentioned technologies, and provides an adaptive algorithm pluggable distributed cluster load forecasting method, which makes up for the problems of non-adaptive selection of algorithms and lack of flexibility existing in existing forecasting methods; Accordingly, a distributed cluster management system is provided

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
  • Cluster load prediction method and distributed cluster management system
  • Cluster load prediction method and distributed cluster management system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0063] Such as figure 1 A cluster load forecasting method shown includes steps:

[0064] S1 sets up the prediction algorithm storage unit and algorithm pool in the system.

[0065] S2 pre-stores the prediction algorithm and its related parameter information that can predict the cluster load in the algorithm pool, and updates the prediction algorithm and its related parameter information in the algorithm pool in real time.

[0066] S3 receives user forecast demand information written externally, and the user forecast demand information includes forecast accuracy information required by the user, forecast rule type information, forecast period information, and forecast algorithm characteristic information; the received user forecast demand information is based on Key-value pairs are stored in the configuration file of the system; the received user forecast demand information is parsed and stored, including steps:

[0067] S301 reads all user forecast demand information one by ...

Embodiment 2

[0081] The specific steps of this embodiment are consistent with Embodiment 1, but in step S3, the received user forecast demand information is stored in the configuration file of the system in XML format, so the received user forecast demand information is analyzed And when storing, specifically include steps:

[0082] Analyze the data root node of the user's forecast demand information, store the root node and its attribute information in the system, and then analyze the child nodes and their attribute information of the user's forecast demand information in a circular traversal manner, and store them in the system;

[0083] By analogy, until all node information of user forecast demand information has been parsed.

[0084] In Embodiment 1 and Embodiment 2, the algorithm pool can be specifically a file in the specified system directory or a database selected when the amount of algorithm information is relatively large, that is, the prediction algorithm and related parameter ...

Embodiment 3

[0103] Such as figure 2 A distributed cluster management system shown includes a cluster scheduling module 1, a load monitoring module 2, a load forecasting module 3 and a decision implementation module 4, and the load forecasting module 3 further includes a configuration file 301, an algorithm controller 302 and an algorithm executor 303.

[0104] The cluster scheduling module 1 stores the externally input user forecast demand information in the configuration file 301, and the specific storage form can be in the form of key-value pairs or in XML format during the specific implementation process.

[0105] Algorithm controller 302, including an algorithm pool 320 that stores forecasting algorithms and related parameter information capable of predicting cluster loads; algorithm controller 302 updates the forecasting algorithms and related parameter information in the algorithm pool 320 in real time, and parses configuration files User forecast demand information in 301; the an...

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 invention relates to a cluster load prediction method and a distributed cluster management system. The cluster load prediction method includes steps of receiving user prediction demand information and analyzing the same, and matching the analyzed user prediction demand information and prediction algorithm stored in an algorithm pool; updating the prediction algorithm in the algorithm pool; building a load prediction model by selecting the mostly-matched prediction algorithm and relevant parameters and training the load prediction model by acquiring load data; and predicting cluster load by the trained load prediction model to obtain prediction results. The distributed cluster management system comprises a cluster dispatching module, a load monitoring module, a load prediction module and a policy implementing module, wherein the load prediction module comprises configuration files, an algorithm controller and an algorithm executor. The cluster load prediction method and the distributed cluster management system in the technical scheme are applied to the cluster system and support configurability of prediction demands, and self-adaptive selection and pluggability of the prediction algorithm.

Description

technical field [0001] The invention relates to a distributed cluster system, in particular to a cluster load forecasting method and a corresponding distributed cluster management system. Background technique [0002] With the continuous development of computer technology in today's society, distributed clusters are more and more widely used in different fields. However, as the scale of clusters continues to increase, traditional cluster management methods can no longer meet the needs of large-scale clusters. Therefore, the application of distributed large-scale cluster management and scheduling technology is born, that is, distributed cluster management systems. According to whether the system predicts the cluster load, its workflow can be divided into two types: [0003] In the first type, the system does not predict the cluster load: the cluster scheduling module directly reads the load data from the load monitoring module, then analyzes the data, makes a scheduling deci...

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): H04L29/08
Inventor 王总辉张涛王云霄陈建海陈文智
Owner ZHEJIANG UNIV