Big Data Analytics
Jon Einkauf
Senior Product Manager, Amazon Elastic MapReduce
1. Introducing Big Data
2. From data to actionable information
3. Analytics and Cloud Computing
Overview
Introducing Big Data
1
Generation
Collection & storage
Analytics & computation
Collaboration & sharing
The cost of data generation
is falling
Generation
Collection & storage
Analytics & computation
Collaboration & sharing
Lower cost,
higher throughput
Generation
Collection & storage
Analytics & computation
Collaboration & sharing
Lower cost,
higher throughput
Highly
constrained
Gartner: User Survey Analysis: Key Trends Shaping the Future of Data Center Infrastructure
Through 2011
IDC: Worldwide Business Analytics Software 2012–2016 Forecast and 2011 Vendor Shares
Generated data
Available for analysis
Data volume
Gartner: User Survey Analysis: Key Trends Shaping the Future of Data Center Infrastructure Through 2011
IDC: Worldwide Business Analytics Software 2012–2016 Forecast and 2011 Vendor Shares
Elastic and highly scalable
No upfront capital expense
Only pay for what you use +
+
Available on-demand
+
= Remove
constraints
Generation
Collection & storage
Analytics & computation
Collaboration & sharing
Lower cost,
higher throughput
Highly
constrained
Generation
Collection & storage
Analytics & computation
Collaboration & sharing
Accelerated
Technologies and techniques for
working productively with data,
at any scale.
Big Data
From data to
actionable information
2
“Who buys video games?”
3.5 billion records
13 TB of click stream logs
71 million unique cookies
Per day:
500% return on ad spend
17,000% reduction in procurement time
Results:
“Who is using our
service?”
Identified early mobile usage
Invested heavily in mobile development
Finding signal in the noise of logs
9,432,061 unique mobile devices
used the Yelp mobile app.
4 million+ calls. 5 million+ directions.
In January 2013
Open web index.
3.4 billion records.
Available to all.
Full parse for impact of
social networks
300 lines of Ruby code.
14 hours.
$100.
You Are What You Tweet: Analyzing Twitter for Public Health. M. J. Paul and M. Dredze, 2011
Tweeting about Flu
Analytics and
Cloud Computing
3
Generation
Collection & storage
Analytics & computation
Collaboration & sharing
Generation
Collection & storage
Analytics & computation
Collaboration & sharing
S3, Glacier,
Storage Gateway,
DynamoDB,
Redshift, RDS,
HBase
Generation
Collection & storage
Analytics & computation
Collaboration & sharing
EC2 &
Elastic MapReduce
Generation
Collection & storage
Analytics & computation
Collaboration & sharing
EC2 & S3,
CloudFormation,
Elastic MapReduce,
RDS, DynamoDB, Redshift
Generation
Collection & storage
Analytics & computation
Collaboration & sharing
EC2 & S3,
CloudFormation,
Elastic MapReduce,
RDS, DynamoDB, Redshift
EC2 &
Elastic MapReduce
S3, Glacier,
Storage Gateway,
DynamoDB,
Redshift, RDS,
HBase AWS Data Pipeline
Elastic MapReduce
How does it work?
EMR
EMR Cluster S3
1. Put the data into S3 (or HDFS)
3. Get the results
2. Launch your cluster. Choose: • Hadoop distribution • How many nodes • Node type (hi-CPU,
hi-memory, etc.) • Hadoop apps (Hive,
Pig, HBase)
EMR
EMR Cluster
How does it work?
S3
You can easily resize the cluster
EMR
EMR Cluster
How does it work?
S3
Use Spot nodes to save time
and money
EMR
EMR Cluster
How does it work?
S3
Launch parallel clusters against the same data source (tune for the
workload)
How does it work?
EMR Cluster S3
When the work is complete, you can terminate the cluster
(and stop paying)
EMR Cluster
How does it work?
You can store everything in HDFS
(local disk)
High Storage nodes = 48 TB/node
EMR Cluster
How does it work?
Launch in a Virtual Private Cloud for
extra security
Thousands of Customers, 5+ Million Clusters
Give it a try.
Cost to run a 100-node EMR cluster:
£4.90 / hour
AWS Data Pipeline
Data-intensive orchestration and automation
Reliable and scheduled
Easy to use, drag and drop
Execution and retry logic
Map data dependencies
Create and manage temporary compute
resources
Anatomy of a pipeline
Additional checks and notifications
Arbitrarily complex pipelines
Thanks. [email protected]
To Learn More:
aws.amazon.com/elasticmapreduce
aws.amazon.com/datapipeline
aws.amazon.com/big-data
Back to the Future
Big Data at Channel 4
Bob Harris
Chief Technology Officer – Channel 4 Television
April 2013
The Disclaimer
<IMHO>
blah blah blah…..
</IMHO>
C4 in the Cloud
• 2008 – Started investigations into Cloud Computing
• 2008 – Launched our first applications on AWS
• 2009 – Entered into an Enterprise Agreement with Amazon for AWS
Rapid growth of AWS based offerings during 2009/2010
• 2011 – AWS established as the default platform of choice for new websites
C4 in the Cloud
C4 in the Cloud
• 2008 – Started investigations into Cloud Computing
• 2008 – Launched our first applications on AWS
• 2009 – Entered into an Enterprise Agreement with Amazon for AWS
Rapid growth of AWS based offerings during 2009/2010
• 2011 – AWS established as the default platform of choice for new websites
• 2012 – Adopted cloud-based analytics
• 2013 – Investigating cloud-based back-up and archiving
Why Big Data?
Business Intelligence at C4
• Well established Business Intelligence capability
• Based on industry standard proprietary products
• Real-time data warehousing
• Comprehensive business reporting
• Excellent internal skills
• Good external skills availability
Big Data at C4
2011
• Embarked on Big Data initiative in 2011
• Ran in-house and cloud-based PoCs
• Selected AWS Elastic Map Reduce
2012
• Ran EMR in parallel with conventional BI stack
• Hive deployed to Data Analysts in 2012
• EMR workflows deployed to production in 2012
2013
• EMR confirmed as primary Big Data platform
• EMR usage growing, focus on automation
• Experimenting with R and Mahout
Big Data at C4 – Elastic MapReduce
• AWS EMR established as our Big Data platform of choice
• Friendly front-end developed to allow Data Analysts to
start/stop clusters and submit/track queries.
Big Data at C4 – Big Data Control Panel
Big Data at C4 – Elastic MapReduce
• AWS EMR established as our Big Data platform of choice
• Friendly front-end developed to allow Data Analysts to
start/stop clusters and submit/track queries.
• Production workflows written predominantly in Python and
Pig
• Fully integrated with our conventional BI stack making
EMR outputs available for reporting
• Experimenting with ADP (AWS Data Pipeline)
• Next steps – MapR and HBase
Personalising the viewer experience
Most popular dramas
Drama
collections
US drama
Single view of the viewer recognising them across devices
and serving relevant content
Big Data – Improving Viewer Experience
Myths or Truths? – It’s all about Perspective!
• Nothing that can’t be done with an RDBMS
• It’s a completely different approach
• It’s really difficult
• It’s immature and lacks good tools
• It’s totally incompatible with you current BI platform
and tools
• It’s difficult to find skilled and experienced staff
Image by Tayrawr Fortune
Elastic MapReduce has provided a cost effective
approach to establishing our Big Data platform
Alan Priestley
EMEA Enterprise Marketing
Intel Corporation
Analysis of Data Can Transform Society
Create new business
models and improve
organizational
processes.
Enhance scientific
understanding, drive
innovation, and
accelerate medical cures.
Increase public safety
and improve
energy efficiency with
smart grids.
Democratizing Analytics gets Value out of Big Data
Unlock Value in
Silicon
Support Open
Platforms
Deliver Software Value
Intel at the Intersection of Big Data
Enabling exascale
computing on massive
data sets
Helping enterprises build open
interoperable clouds
Contributing code and fostering ecosystem
HPC Cloud Open Source
Intel at the Heart of the Cloud
Server
Storage
Network
Scale-Out Platform Optimizations for Big Data
Cost-effective performance
•Intel® Advanced Vector Extension Technology
•Intel® Turbo Boost Technology 2.0
•Intel® Advanced Encryption Standard New
Instructions Technology
66
Intel® Advanced Vector Extensions Technology
• Newest in a long line of
processor instruction
innovations
• Increases floating point
operations per clock up to
2X1 performance
1 : Performance comparison using Linpack benchmark. See backup for configuration details.
For more legal information on performance forecasts go to http://www.intel.com/performance
Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products.
Intel® Turbo Boost Technology 2.0
More Performance Higher turbo speeds maximize
performance for single and
multi-threaded applications
Intel® Advanced Encryption
Standard New Instructions
• Processor assistance for performing AES encryption 7 new instructions
• Makes enabled encryption software faster and stronger
Power of the Platform built by Intel
Richer
user
experiences
4HRS
50% Reduction
10MIN
80% Reduction 50%
Reduction 40% Reduction
TeraSort for
1TB sort
Intel®
Xeon®
Processor
E5 2600
Solid-State
Drive 10G
Ethernet Intel® Apache
Hadoop
Previous
Intel®
Xeon®
Processor
Cloud
Intelligent Systems
Clients
Virtuous Cycle of Data-Driven Experience
Get 600 Hours of free supercomputing
time!
www.powerof60.com
Thank you!