Separating Hadoop Myths from Reality by ROB ANDERSON at Big Data Spain 2013
-
Upload
big-data-spain -
Category
Technology
-
view
106 -
download
0
description
Transcript of Separating Hadoop Myths from Reality by ROB ANDERSON at Big Data Spain 2013
Separating Hadoop Myths from Reality
Rob Anderson
1
The Myths & Reali.es Surrounding Hadoop Rob Anderson VP Systems Engineering
2
Sales
SCM CRM
Public
Web Logs Produc7on
Data Sensor Data Click
Streams Loca7on
Social Media
Billing
Enterprise Data Hub
Hadoop Changes Analy.cs
“Simple algorithms and lots of data trump complex models ”
Halevy, Norvig, and Pereira, Google IEEE Intelligent Systems
3
4
5
Data Warehouse
Volume
Variety
Velocity
6
7
Big Data is hard to move…because it’s BIG
8
What was the genius of Hadoop?
§ Fueling an industry revolu7on by providing infinite capability to store and process big data
§ Expanding analy7cs across data types
§ Compelling economics – 20 to 100X more cost effec7ve than alterna7ves
9
10
Random Wri.ng in MapR S1
S2
S3 S5 S4
S1, S2, S4 S1, S3 S1, S4, S5 S2, S4, S5 S3
Client wri.ng data
CLDB Ask for 64M block
Create cont.
Picks master and 2 replica slaves
Write next chunk to S2
S2, S3, S5
aZach
11
12
MapR Spout
TwiZer
TwiZer API
TwiZerLogger
Storm MapR
Op7onal MapReduce
DFS
13 hZp://www.flickr.com/photos/onemoreshotrog/8085462024/
14
Hadoop Distribu.ons
Hadoop: The Disrup.ve Technology at the Core of Big Data
16
17
The Reality is Architecture MaHers
MapR Data System
Architecture Comparison
HBase
JVM
HDFS
JVM
ext3/ext4
Disks
Other Distribu7ons
Disks
MapR M7
Architecture Results
Results with other distribu.ons
Results with MapR M7
20
Produc.on Success with Hadoop
22
2000+ Nodes Fortune 100 Retailer
23
1000+ Nodes Fortune 100 Financial Services Company
24
25
Produc7on Hadoop in Waste Management
Waste Management Logis.cs
26
Suntory whiskey
27
28
Unique Iden.ty Ini.a.ve, India
30
Thank you Big Data Spain!