A cloud server load forecasting method

Abstract

The invention relates to a cloud server load prediction method. Aiming at the characteristics of load data, the advantages of the cloud model and the Markov chain are fused to propose a cloud server load prediction method. The method follows the historical data sample training method of the cloud model. In the calculation of the degree of membership value of the predicted value, the predicted degree of membership vector is obtained by Markov forecasting. Then according to the predicted degree of membership vector, the predicted value weighted summarization mode is adopted for the cloud module, so as to achieve the organic combination of the cloud model and the Markov chain. Finally, the load prediction of the cloud server is achieved.

Description

technical field [0001] The invention relates to a cloud server load prediction method, which belongs to the technical field of cloud computing applications. Background technique [0002] Cloud computing is a computing model that uses the Internet to access shared resource pools (such as computing facilities, storage devices, applications, etc.) anytime, anywhere, on-demand, and conveniently. Resources, which can reduce hardware maintenance costs for small and medium-sized enterprises. At present, the total number of various data centers in my country is about 430,000, which can accommodate about 5 million servers. In the next five years, my country's demand for data center traffic processing capacity will increase by 7 to 10 times. However, the continuous expansion of the scale of cloud data centers and thousands of computing nodes have also brought about high energy consumption. Based on the average power consumption of a server of 400 watts, the total annual power consump...

AI Technical Summary

Problems solved by technology

The load data obtained from monitoring cannot be fitted with a linear function, and the data has uncertain properties

[0005] (2) Strong volatility

[0007] (1) The prediction accuracy of non-heuristic methods is low, and it is difficult to predict accurately for volatility data;

[0008] (2) The heuristic method has a large processing cost, and the multiple iterations in the method cause a huge computational cost, which cannot be applied to load prediction that requires strong real-time performance;

[0009] (3) The prediction time granularity and target granularity of the method are too large. Some load prediction methods use days as the time unit to predict the overall cloud environment load. The purpose is only to predict the possible data center task volume of the day to confirm the number of servers opened on the day. Little contribution to real-time management of cloud data center, limited energy saving

Images