雲端計算與應用
Cloud Computing and its Applications

課程資訊 Course Info.

• 上課時間： 2010/7/12 (一) ～ 2010/8/20 (五) 13:30 ～ 16:20
• Date and Time: 13:30 to 16:20, from 12 July 2010 to 20 Aug. 2010, every Monday, Wednesday and Friday
• 上課地點： 國家高速網路與計算中心 新竹事業群(300 新竹市科學工業園區研發六路七號) <​地圖> 階梯教室
• Location: No. 7, RD VI Road, Hsin-Chu, Taiwan - National Center for High-performance Computing

課程大綱 Course Outline

時段 Date	分類 Section	課程內容 Topics	主講人 Lectuer	投影片 Slides	實作 Hands-On	補充資料 Notes
7/12	Introduction	Introduction to Parallel, Cluster, Grid Computing and Web Service	東海大學資工系 ​楊朝棟 教授 ​Prof. Chao-Tung Yang	1-1
7/14	Introduction	The Anatomy of cloud computing	東海大學資工系 ​楊朝棟 教授 ​Prof. Chao-Tung Yang	1-2		HW1
7/16	Theory	Distributed File System: Overview	中研院資科所 ​王建民 副研究員 ​Dr. Chien-Min Wang	2-1
7/19	Theory	Distributed File Systems: Issues	中研院資科所 ​王建民 副研究員 ​Dr. Chien-Min Wang	2-2
7/21	Theory	Distributed File Systems: Case Studies	中研院資科所 ​王建民 副研究員 ​Dr. Chien-Min Wang	2-3
7/23	Theory	Distributed File Systems for Clouds	中研院資科所 ​王建民 副研究員 ​Dr. Chien-Min Wang	2-4		HW2-1 HW2-2
7/26	Theory	Workload Model, Job Scheduling, and Resource Allocation	台中教育大學資科系 ​黃國展 助理教授 Prof. Kuo-Chan Huang	3-1
7/28	Theory	Workload Model, Job Scheduling, and Resource Allocation	台中教育大學資科系 ​黃國展 助理教授 Prof. Kuo-Chan Huang	3-2
7/30	Theory	Workload Model, Job Scheduling, and Resource Allocation	台中教育大學資科系 ​黃國展 助理教授 Prof. Kuo-Chan Huang	3-3		Exam 3
8/02	Skills	- Build Your Private Cloud with Open Source - Introduction to Hadoop : History and Terminology	國網中心軟體技術組 王耀聰 副研究員 ​Jazz Yao-Tsung Wang	4-1 4-2		- 四頁列印版/4 pages
8/04	Hands-on	- Introduction to HDFS - Hadoop Installation: presudo mode	國網中心軟體技術組 王耀聰 副研究員 ​Jazz Yao-Tsung Wang	4-3	Lab1 Lab2	@ PC Room C
8/06	Hands-on	- Introduction to MapReduce - MapReduce Programing 101 - Introduction to Clonezilla	國網中心軟體技術組 王耀聰 副研究員 ​Jazz Yao-Tsung Wang	4-4 ​4-5	Lab3 Lab4	@ PC Room B
8/09	Hands-on	- Hadoop Fully Distributed Mode - Hadoop Deployment Tool : SmartFog and DRBL	國網中心軟體技術組 王耀聰 副研究員 ​Jazz Yao-Tsung Wang	4-6	Lab5	@ PC Room B
8/11	Skills	- Hadoop Ecosystem : HBase, Pig and Streaming	國網中心軟體技術組 王耀聰 副研究員 ​Jazz Yao-Tsung Wang	4-7	Demo1	HW3
8/13	Application	- Opportunities and challenges for large-scale data collections - Introduction to photo/video representations	台大資工系 ​徐宏民 助理教授 ​Prof. Winston H. Hsu
8/16	Application	- Multimodal image similarities - Graph construction, clustering, and learning	台大資工系 ​徐宏民 助理教授 ​Prof. Winston H. Hsu
8/18	Application	- Data warehousing - Clustering algorithms	台大資工系 ​徐宏民 助理教授 ​Prof. Winston H. Hsu
8/20	Application	- Visual semantic analysis - Machine learning algorithms	台大資工系 ​徐宏民 助理教授 ​Prof. Winston H. Hsu			HW4

HW1

HW2-1

HW2-2

HW3

• Due: 2010/08/20 (Fri.)
• Try to run hadoop streaming job on ​http://hadoop.nchc.org.tw
  • Mapper: /bin/cat
  • Reducer: /usr/bin/wc
  • Input: /etc/hadoop/conf
  • Output: /user/${id}/output
• Run with extra parameter with -numReduceTasks
• Please take screenshot of both results and compare the results. Compare the time of 1 reducer with 8 reducers.
• Send a report of this homework to jazz_AT_nchc_DOT_org_DOT_tw

HW4

• The goal for the assignment is to exercise the MapReduce implementation for constructing inverted file indexing from the small set of (nine) documents. The document id is at the first number.
• data could be download from here.
• The due is on August 27, Friday.
• Please send the source code and the output to the lecturer winston AT csie DOT ntu DOT edu DOT tw