基於雲端運算公有雲之資源利用率分析與預測框架

Abstract

在雲端基礎設施服務(Infrasructure as A Service )裡，使用者可以向雲端服務的提供者承租虛擬機器來執行它們的程式，但是虛擬機器有多種的規格，使用者對於自己的程式應該花多少個虛擬機器來執行毫無概念。因此，我們提出了一個可以預估資源的服務框架，此服務框架能夠告訴使用者對於即將執行的程式需要最少需要多少個虛擬機器執行才能在特定的時間內完成，為了驗證此預估資源的服務框架可行性，我們還做了一些範例的研究，由於Frequent Pattern Growth (FP-growth)、K-means、Particle Swarm Optimization (PSO)的輸入資料皆相當大量，運算複雜，因此我們將此三種演算法修改成平行化的執行方式，並運行在此框架底下進行資源預估，實驗結果顯示我們此框架可以成功地針對某一個新來的工作預估出在特定時間內完成所需要的虛擬機器個數。另外，從範例演算法中得知此框架不僅可運行輸入資料量小的工作，亦適用於輸入資料很大量的工作，此框架相當彈性，此彈性化以及實用的框架讓使用者在租用虛擬機器時不會租太多虛擬機器而浪費錢，也讓雲端服務的提供者能夠更方便地管理系統內的虛擬機器。

In cloud computing environments, users can rent virtual machines (VMs) from cloud providers to execute their programs or provide network services. While using this kind of cloud service, one of the biggest problems for the users is that how many VMs are needed to complete the jobs without spending too much money and time. In this paper, we propose a resource prediction framework (RPF) which can help users rent the minimum number of virtual machines and complete their jobs within a user specified time constraint. In order to verify the feasibility of RPF, we have done 3 case studies, parallel frequent pattern growth (FP-Growth), parallel K-means and Particle Swarm Optimization (PSO), on the proposed framework. FP-growth, K-means and PSO are data intensive algorithms. These algorithms may be executed repeatedly with different execution parameters to find the optimal results. When evaluating RPF by these algorithms in cloud environments, we have to modify them to parallel versions. The evaluation results indicate that RPF can successfully obtain the minimum number of VMs with acceptable errors. According to the results of case studies, the proposed RPF can be adopted by data intensive jobs, which is flexible and useful for users and cloud system providers.