近年來由於雲端運算概念的興起,網際網路服務的應用,已透過雲端運算擴大其服務層面的廣度及服務內容的深度,許多科技公司開始建立自己的雲端服務,其中像是Google Apps及Apple的多媒體服務,但建置雲端服務的數據中心也產生了龐大的電能耗費及二氧化碳的排放,並且大多數的數據中心資源並沒有得到充分的利用。因此要如何有效利用伺服器的資源及降低能源消耗成了一個重要的議題。 本論文提出一個改良式最佳遞減裝箱配置(MBFDP, Modified Best Fit Decreasing Packing)方法並整合需求預測模型,針對雲端伺服器動態節能的改進方法,以降低伺服器的能源耗費及提高伺服器的資源利用率,而其中關鍵的技術如虛擬化技術等也將在本文中論及。
The applications of the Internet service have been extended rapidly since the raising of the cloud computing concepts. However, constructing a data center of the cloud computing service also has been creating the problem of power consuming and carbon dioxide emission. Since the resources of data center have not been used effectively, the decreasing of power consumption and increasing efficiency of server’s resources are important issues in the cloud computing environment. This thesis presents an improving method, MBFDP (Modified Best Fit Decreasing Packing) , to integrate servers for decreasing power consuming and to enhance resources usage of the server resources. Some key techniques, such as virtualization and packing problem, will also be discussed.