美國國家標準與技術研究院（NIST）將雲端定義為：「雲端運算是一種模式，能方便且隨需求應變地透過連網存取廣大的共享運算資源（如網路、伺服器、儲存、應用程式、服務等），並可透過最少的管理工作及服務供應者互動，快速提供各項服務。」根據Gartner諮詢公司的分析，雲端運算是2015年對企業組織而言最重要的策略科技趨勢的前十大之一。所謂的策略科技(strategic technology)，根據Gartner定義，指的是可能在未來三年對企業組織帶來重大影響的技術。各個企業、組織與學校也都跟隨著雲端的潮流，建立大規模的雲端運算叢集取代一人一電腦的情形。雖然虛擬化可以減少添購硬體設備的，但是也衍生出了兩個問題-能源的消耗與閒置資源的浪費。所以我們提出了兩個方法:1.動態調配資源方法2.節電方法。如何有效節省並利用虛擬機於低負載時的閒置資源，與如何節省伺服器的能源消耗，是我們在本篇論文裡必須面對與解決的兩大問題。為了達到我們的目標，我們實現一個以雲端軟體OpenStack為基礎設施的平台並使用動態調配資源方法與節電方法來達到節省能源的目的。我們也將會利用PDU紀錄的耗電量來證明我們提出的方法是有用而且真的可以節省能源。 The US National Institute of Standards and Technology (NIST) defines cloud computing as “a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction.” According to analysis by Gartner, Inc., cloud computing is one of the top 10 strategic technology for most organizations in 2015. Gartner defines a strategic technology as one with the potential for significant impact on the organization in the next three years. Companies, organizations and academic institutions are following the cloud computing trend; the establishment of large-scale cloud computing clusters avoids the need to provide one person with one computer. Even though virtualization can reduce the cost of hardware equipment, but it still faces with two problems: energy consumption and the waste of the idle resources. To solve these two problems, we propose two algorithms, i.e., dynamic resource allocation and energy saving. In order to implement these two algorithms with live migration of virtual machines, we first build an infrastructure platform based on cloud software – OpenStack. Next, dynamic resource allocation and energy saving algorithms are designed and implemented. Finally, we use the Power Distribution Unit (PDU) to monitor system status and record power consumption; the real time status monitoring data verify that the proposed algorithms are efficient in energy saving and idle resource planning.
