R公司之高效率資訊中心建置與節能探討

Study on company R: High Efficient & Energy Saving Information Centre

徐凡耘(Fen-yun Hsu)

指導教授 ： 范錚強

國立中央大學/管理學院/資訊管理學系碩士在職專班/碩士(2011年)

摘要

在過去，許多傳統的資訊中心強調的是可用度，並未考量到整體規劃與未來的需求，總是在既有的空間裡不斷進行設備擴充。面對著不斷成長的資訊需求與不斷的引進或開發新的資訊應用系統，當然伺服器部署數量也就越來越多，服務的項目也越來越複雜，無論在種類上或數量上皆呈現快速增長的局勢。而支援這些資訊應用系統運作的資訊基礎架構，例如伺服器與儲存體，更是以倍數在成長，使得許多資訊中心普遍皆呈現「機滿為患」的現象。資訊中心面臨了諸如基礎設施之電力、空調、網路等系統瓶頸之外，也增加了許多管理面與機器耗電等各式有形與無形的成本。
本研究的目的是提出一個設計新的節能且高效率資訊中心的嘗試。本研究主要使用TileFlow™ CFD電腦模擬分析軟體，首先分析舊有資訊中心與模擬分析之差異性，找出最適合的參數。接下來，再將此參數套用於新的資訊中心。希望藉此能打造一個高效率資訊中心之前，有一個概估狀況的模擬，提供設備配置上的一個參考。本研究同時提出其他必須的配套設計，來配合新的資訊中心。
高效率資訊中心除了強調其效能的最大化與成本的最小化，在打造一個高效率資訊中心同時，除了強調其靈活性 (Agility)、自動化 (Automation)、安全性 (Security)與完善的服務層級 (Service Level)之外，納入新技術應用也是重點項目。由於資訊中心管理沒有法令依據可以參考，所以按照業界相關標準如國際標準規範 (TIA-942, Uptime Institute, TGG and ASHREA) 做為設計基礎，才能提供合理的評估標準，藉此增加企業的競爭力。

關鍵字

機櫃冷卻指數 ； 電源使用效率 ； 流體力學計算 ； 節能資訊中心 ； 電源分配器 ； 被動式煙囪型機櫃冷卻

並列摘要

In traditional data centers, the emphasis has always been placed on the availability of physical spaces without planning for future technology integrations and requirement growths. A common scenario is that, in a limited space, more and more servers and equipment were installed as the need of information processing increased, and the new application systems developed. As a result, the services offered have grown rapidly and become more perplexed both in numbers and types.
To support the growth in requirements, equipment, such as servers and data storage devices, has multiplied in numbers. Therefore, many data centers experience the so called “machine-infested” phenomenon. These data centers have not only faced the basic operation bottlenecks such as electricity, air-conditioning, and networking, but have also faced with the increasing tangible and intangible costs in energy waste and management.
This study attempted to design a new energy saving and efficient data center. It adopted TileFlow™ CFD, a software tool for stimulating cooling performance of data centers. First of all, it analyzed the environment in the current center and compares with the scenario set in the stimulation to find the closet parameters. Then, it applied this set of parameters to the new to-be-built data center. It then calculated the efficiency of data center performance with the stimulated hypothesis and provided the equipment setup parameters. This study also proposed coupling features for new data center.
A high efficient data center focuses not only on maximizing its performance but also on minimizing its cost. When designing a high efficient data center, the key factors include but not limited to agility, automation, security, and service level. The capability of integrating new technology applications is also an important component. Since there is currently no formal law regulating data center management, the common practice in the industry is to follow related international standards as a basis when designing data centers. The general adopted international standards include TIA-942, Uptime Institute, TGG, and ASHERA. These guidelines provide a set of sound measurements to business in increasing its competitiveness in the industry.

並列關鍵字

Passive cooling rack ； Green data center ； PDU ； CFD ； PUE ； RCI

參考文獻

〔1〕 王翔勇，(2009)，資訊中心耗能監測與模擬分析–以某商業金融中心為例，碩士論文，國立台北科技大學，能源與冷凍空調工程系。

〔3〕 張銘陽，(2009)，網路資訊中心空調節能設計之模擬分析研究，碩士論文，國立台北科技大學，能源與冷凍空調工程系。

〔4〕 Aperture, Inc. (2009). Computational Fluid Dynamics Modeling for Operation Data Center, Aperture, Inc.

〔7〕 Info-Tech Research Group (2007). Top 10 Energy-Saving Tips for a Greener Data Center, Info-Tech Research Group.

〔10〕 Ian Seaton (2008). A Path to Higher Density and Lower Cost, CPI, Inc.