Software Design And Implementation Of A Recession Forecast

Now, software aging is a very common phenomenon. A number of measurements and studies of software system indicate that the main reason is exhaustion of operation system. This paper aims to provide data for the application, formulate appropriate recovery strategy and enhance the reliability of the system by using the software performance monitoring and recovery tools.The main contents of this paper are as follows:(1) Study on related technology to find out appropriate acquisition technology, and use the/proc virtual file system to get resource consumption data. Auto-Regressive Moving-Average (ARMA) model is adopted in order to build the corresponding forecast model. Forecasting object is Holder exponent which was analyzed by fractal analysis method not raw data.(2) Analyze recession prediction system requirements, configure collecting environment, adopt strategies such as data acquisition, data consolidation, data transmission, and carry on the qualitative analysis on performance parameters of the system characteristics, and the quantitative prediction of Holder index sequence; Design system functional module for detecting node to get information about resources consumption; analyze the collected data, which will be stored by the database. Study the fractal characteristics of resource data to get the multi-dimensional Holder index time series, as a historical observation, make a research on its self correlation and partial correlation, and establish the right model on the prediction of time series.(3) Set up the proper acquisition parameters according to actual needs, understand the Socket communication, establish communication rules between a client and server, and achieve recession prediction module by predicting the multifractal spectrum and the multi-dimensional index.