Feature-based load selecting method and system

Abstract

The invention applies to the field of algorithms, and provides a feature-based load selecting method and system. The method comprises the steps of preprocessing feature data to be processed; classifying the feature data to be processed through the feature clustering algorithm; acquiring representative element of each class; selecting high-accuracy load corresponding to the feature according to the mutual information and the representative elements. According to the method, the feature data to be processed are preprocessed and then classified through the feature clustering algorithm to obtain the representative element of each class, and then the high-accuracy load corresponding to the feature can be selected according to the mutual information value and the representative elements; the method and the system are high in efficiency and can increase the load selection efficiency.

Description

technical field

[0001] The invention belongs to the field of algorithms, and in particular relates to a method and system for selecting loads according to characteristics. Background technique

[0002] Whether it is a traditional physical machine or a virtual cluster in cloud computing, it is very important for system optimization. In order to adapt to different application requirements, different optimization methods will be adopted for system optimization. In this case, it is first necessary to classify the loads of physical machines or virtual machines, and adopt different optimization methods according to whether they are CPU-intensive, memory-intensive, IO-intensive, or network-intensive to improve efficiency.

[0003] The load classification method is the premise of system optimization, and its efficiency directly affects the efficiency of system optimization. In the process of load classification, accuracy and efficiency are mutually restrictive factors, and usually...

Images