Adaptive scaling control system and method for web application in cloud computing platform

Abstract

The invention discloses a self-adapting flexible control system of Web application in a cloud computing platform and a method of the self-adapting flexible control system, which are used to dynamically adjust computing resources according to load change. The system comprises a performance monitor, a load database, a performance model computing module, an optimization controller and an automatic configuring module. The method comprises the steps that: firstly, the performance monitor constructs a layered queue network model according to structure of the Web application and processing procedure of requests; the web application is deployed in a real cloud computing platform and a record label is inserted into each layer of component of the web application, so as to record the actual execution time of each request at each resource of each component, thereby obtaining parameters needed by a web application performance model in the performance model computing module; when application load changes, the optimization controller computes the performance of the application of each resource configuration scheme via a heuristic search algorithm, finds the configuration scheme which needs the minimal cost and also can meet the requirement on Qos (quality of service), and then takes the configuration scheme as the optimal configuration scheme; and finally, the automatic configuring module readjusts resources needed by each component of the application.

Description

technical field [0001] The present invention relates to the effective management of resources in a cloud platform, and in particular to an adaptive scaling control system for multi-layer web applications and a working method thereof. The present invention supports dynamic adjustment of computing resources as load changes. Background technique [0002] Using an adaptive scaling control system can help service providers dynamically schedule resources according to load requirements, so as to meet users' QoS (Quality of Service, quality of service) requirements with the least resources. Compared with the traditional access control technology, the adaptive scaling control system will not determine the maximum number of users that the application can accommodate, nor will it reject the customer's request because of too many users, and it can guarantee the overall service quality, and at the same time No additional resources are used. The current adaptive scaling control technolog...

AI Technical Summary

Problems solved by technology

The current adaptive scaling control technology in the cloud platform is divided into two types: (1) The method based on the virtual machine: this method does not consider the specific application inside the virtual machine, and formulates a specific scaling strategy, based on the current resource utilization of the virtual machine. Adjust the allocation of resources, its defect is: From the perspective of the entire application, the performance improvement of a single component virtual machine may transfer the performance bottleneck to other component virtual machines

(2) Application-based method: The current method often assumes that the application has only a single transaction type, and the number of requests that each virtual machine can handle is a fixed value, so as to scale according to the total load. In fact, this method is difficult to expand to multiple Automatic scaling of transactional web applications

At the same time, these two methods currently do not take into account the situation of multiple QoS

Images