Study on Operation and Management in Server rooms using Machine Learning Prediction Model
-
- SASAKURA Kosuke
- NTT FACILITIES INC.
-
- KOMATSU Masahiro
- NTT FACILITIES INC.
-
- AOKI Takeshi
- NTT FACILITIES INC.
-
- WATANABE Takeshi
- NTT FACILITIES INC.
Bibliographic Information
- Other Title
-
- 機械学習モデルを活用したサーバルームにおける温度環境の管理手法に関する研究
- Part 1-Proposal on Prediction Model of Rack Intake Temperature
- 第1報−ラック吸気温度の予測モデルに関する提案
Abstract
<p>In recent years, data centers (DCs) have become increasingly important. Accordingly, highly efficient and reliable operation and management of DCs are required. The conventional layout design of racks and ICT equipment in a server room considers the current space and capacity of the power and cooling status, but it does not consider the temperature or energy performance after environmental changes such as rack removal or addition. In contrast, to manage the temperature in the server rooms, the DC users usually require the DC operator to comply with a service level agreement that stipulates that the rack intake temperature should be kept below a certain level. However, it is not possible to know the temperature after environmental changes. To address this problem, we constructed some models to predict the rack intake air temperatures in the server rooms by using information related to the ICT equipment, electric power equipment, and air conditioning equipment. Through this study, we propose a method to predict the temperature after environmental changes by using machine learning techniques. We elucidate the characteristics of the proposed method and the influence of each parameter on accuracy, and report the results of its verification and effectiveness in multiple verification rooms. </p>
Journal
-
- Transactions of the Society of Heating,Air-conditioning and Sanitary Engineers of Japan
-
Transactions of the Society of Heating,Air-conditioning and Sanitary Engineers of Japan 45 (278), 1-8, 2020-05-05
The Society of Heating, Air-Conditioning & Sanitary Engineers of Japan
- Tweet
Keywords
Details 詳細情報について
-
- CRID
- 1390006440921383424
-
- NII Article ID
- 130008034659
-
- ISSN
- 24240486
- 0385275X
-
- Text Lang
- ja
-
- Data Source
-
- JaLC
- CiNii Articles
-
- Abstract License Flag
- Disallowed