2012/8/1 夏合宿 発表資料

15
合合 合合 20012/8/1 In 合合合合 合合 合

Upload: keiichi-maeda

Post on 01-Nov-2014

365 views

Category:

Technology


4 download

DESCRIPTION

 

TRANSCRIPT

Page 1: 2012/8/1 夏合宿 発表資料

夏合宿 発表20012/8/1

In 山代温泉

前田 経一

Page 2: 2012/8/1 夏合宿 発表資料

What is “The BIG Switch”

MapReduce and Functional Programming

Topics

Page 3: 2012/8/1 夏合宿 発表資料

Nicholas Carr “The BIG SWITCH : Rewiring the World, from Edison to Google”

Buzzword “Cloud” spread by the book.

Page 4: 2012/8/1 夏合宿 発表資料

Electricity From Edison : private power generation To Utility : central electricity generating

systemComputing as well From Private computing To Utility computing

From Edison to Google

Page 5: 2012/8/1 夏合宿 発表資料

Why “Private to Utility”

Page 6: 2012/8/1 夏合宿 発表資料

Urbanization

Page 7: 2012/8/1 夏合宿 発表資料

Independent vs Dependent

Do It Yourself vs Utility

Generalist vs Specialist

Country vs City

Page 8: 2012/8/1 夏合宿 発表資料

Google

Page 9: 2012/8/1 夏合宿 発表資料

MapReduce

Page 10: 2012/8/1 夏合宿 発表資料

Hadoop

Page 11: 2012/8/1 夏合宿 発表資料

Neumann architecture Program & Data

Data → Program Program → Data

Mobile Agent

Data Locality

Page 12: 2012/8/1 夏合宿 発表資料

“The problem is simple: while the storage capacities of hard drives have increased massively over the years, access speeds—the rate at which data can be read from drives— have not kept up. “

Random Access > Sequential Access

Data access speed

Page 13: 2012/8/1 夏合宿 発表資料

map(String key, String value): // key: document name // value: document contents for each word w in value: EmitIntermediate(w, "1");

Map Function

Page 14: 2012/8/1 夏合宿 発表資料

reduce(String key, Iterator values): // key: a word // values: a list of counts int result = 0; for each v in values: result += ParseInt(v); Emit(AsString(result));

Reduce Function

Page 15: 2012/8/1 夏合宿 発表資料

Map and Reduce concepts came from Functional Programming Language.

Reduce Function must have associativity becouse of non-deterministic.

DAG + Operator

MapReduce and Functional Programming