Research on service delivery problem of multilayer application in cloud

Research on service delivery problem of multilayer application in cloud

Shandong University Zhang Sweets

This paper presents a challenge to the problem of multi-tier application service delivery in the cloud, this paper mainly studies: 1, constructs the on-line monitoring structure, carries on the monitoring to each layer load distribution of multi-layer application, and proposes the forecast method based on the autoregressive model to predict the application load, and solves the dependency complexity problem between layers of multilayer application. , the service characteristics of each layer are monitored, which solves the problem of the complexity of service characteristics among multilayer applications. 2, the application of queuing theory to the deployment of multiple layers of resources to model the application, it solves the complexity problem of resource processing ability by solving the resource's ability to deal with the application of each layer, puts forward a multi-objective optimization algorithm based on performance-cost equilibrium, and uses Pareto optimal idea to obtain the service delivery strategy of better service quality and resource cost. Solve the combinatorial complexity problem of multi layer mixed resources. In this paper, the multilayer application datum test Rubis is used to verify the method presented in this paper through a large number of experimental data. Based on the Rubis running data, the load is forecasted, and the actual running data is compared with the load predicted by the prediction method presented in this paper. The experimental results show that the load forecasting method and the actual load error are smaller and the prediction method has better performance. On the other hand, through the experiment, this paper compares the service delivery strategy and the stochastic strategy and greedy strategy from the overall performance and resource cost of the service delivery scheme. The experimental results show that, compared with the similar service delivery strategy, this paper presents a performance-cost-balanced multi-objective optimization service delivery strategy, which has better comprehensive performances, higher service quality and more resource cost. The research results of this paper provide a basis for improving the utilization of cloud basic resources and accurate service delivery, and have high practical value and wide application prospect.

Research on service delivery problem of multilayer application in cloud