Vous êtes sur la page 1sur 16

BIG DATA

ARMAN SHAIKH
MMS SYSTEMS
M020
1

FLOW OF BIG DATA

Website

Social Media
Billing
ERP
CRM

RFID

Network Switches
2

CHARACTERISTICS OF BIG
DATA

Volume

Velocity

Variety

Data
Quantity

Data
Speed

Data
Type
3

DATA IN 2013
United States

China

India

Western Europe

32%

19%

Rest of the World

32%

13%

4%

PREDICTION OF DATA IN 2023


United States

China

India

Western Europe

Rest of the World

23%
35%

21%

15%

6%

FACTS
By 2016, the cumulative size of all words data center
is expected to 16,000 acres

Only 33 % of the World data can be analyzed


The amount actually analyzed is only 0.5 %
6

HADOOP
Open-source software framework from Apache
Inspired by
Google MapReduce
GFS (Google File System)

HDFS
Map/Reduce

EDITIONS OF HADOOP
Enterprise Edition

Enterprise class

Licensed

Application accelerators
Pre-built applications
Text analytics
Spreadsheet-style tool
RDBMS, warehouse connectivity
Basic Edition
Administrative tools, security
Free download
Eclipse development tools
Performance enhancements
Integrated install
Online Info Center
Apache Big Data Univ.
Hadoop

Breadth of capabilities

HADOOP = MAPREDUCE + HDFS

10

MAP REDUCE
INPUT
DATA

Allows massive

MAP

scalability across
Hadoop servers

MAP

MAP

SHUFFLE

REDUCE

REDUCE

RESULT

11

WHAT ID HADOOP DISTRIBUTED FILE


SYSTEM (HDFS)?
Data in Hadoop is Broken Down into blocks
Distributed throughout the cluster i.e. the Servers
In this way map and reduce functions can be executed which provides the
scalability for Big Data Processing

HDFS tolerates disk Failures by storing multiple copies of each data block
on different servers

Individually blocks are also separated and stored


12

WHAT IS NOSQL ?

13

APPLICATIONS OF BIG DATA


ANALYTICS

Health care
Telecom
Traffic control
Trade analytics
Manufacturing

14

LEADING TECHNOLOGY VENDORS

IBM NETEZZA
ORACLE EXADATA
FRACTAL CONCORDIA
SAS ADVANCED ANALYTICS
15

16

Vous aimerez peut-être aussi