Jarut N.
Cisco Systems
Data. Does it Matter?
AutomotiveAuto sensors reporting location, problems
CommunicationsLocation-based advertising
ConsumerPackaged GoodsSentiment analysis of what’s hot, problems
Financial ServicesRisk and portfolio analysis New products
Education and ResearchExperiment sensor analysis
Competitive success depends on Data and Analytics
Data & Analytics are Top of Mind in Every IndustryHigh Technology/ Industrial Mfg.Mfg. quality warranty analysis
Life SciencesClinical trialsGenomics
Media/EntertainmentViewers/advertising effectiveness
On-Line Services/ Social MediaPeople and career matchingWeb-site optimization
Health CarePatient sensors, monitoring, EHRsQuality of care
Oil and GasDrilling exploration sensor analysis
RetailConsumer sentimentOptimized marketing
Consumer Travel andTransportationSensor analysis for optimaltraffic flowsCustomer sentiment
UtilitiesSmart-meter analysis for network capacity,
Law Enforcement and DefenseThreat analysis - social media monitoring, photo analysis
Data Management and Agility
• Analyzed service sales opportunities in one-tenth the time, at one-tenth the cost
• $40 million in incremental service bookings in the current fiscal year as a result of this initiative
• Implemented a multitenant enterprise platform while delivering immediate business value
CHALLENGES
SOLUTION
INSIGHTS AND IMPACT• Unlock the business value of large data sets• Provide SLAs for internal customers using Big
Data analytics services• Support multiple internal users on same platform
• Implemented enterprise Hadoop platform on Cisco Integrated Infrastructure for Big Data
• Automated job scheduling and process orchestration using Cisco Tidal Enterprise Scheduler
3.8 PETABYES Under 1 management domain!
Big Data/AnalyticsData Management
Data Virtualization / Integration Analytics / Business Intelligence
A Holistic View of a Big Data System
ETL
Real-TimeStreams
Unstructured Data (HDFS)
Real-Time StructuredDatabase
(hBase, Gemfire, Cassandra)
Big SQL(Greenplum,AsterData,
Etc…)
BatchProcessing
Real-TimeProcessing
Analytics
§ Big Data is not just a technology. It is a business strategy for capitalizing on information resources.
How Do We Define Big Data?
Volume
of Tweets created daily.
12+ terabytes
More than just Data Warehousing: Big Data’s Value Is in the Analytics
Volume
of Tweets created daily.
12+ terabytes
Variety
of different types of data.
100s
More than just Data Warehousing: Big Data’s Value Is in the Analytics
Volume
of Tweets created daily.
12+ terabytes
Variety
of different types of data.
100s
trade eventsper second.
5+ million
Velocity
More than just Data Warehousing: Big Data’s Value Is in the Analytics
Volume
of Tweets created daily.
12+ terabytes
Variety
of different types of data.
100sVeracity
decision makers trust their information.
Only 1 in 3
More than just Data Warehousing: Big Data’s Value Is in the Analytics
trade eventsper second.
5+ million
Velocity
Volume
of Tweets created daily.
12+ terabytes
Variety
of different types of data.
100sVeracity
decision makers trust their information.
Only 1 in 3
trade eventsper second.
5+ million
Velocity
More than just Data Warehousing: Big Data’s Value Is in the Analytics
It is all about better analytics on a broader spectrum ofdata, and therefore represents an opportunity to create
even more differentiation among industry peers.
Data Warehousing
Data warehousing is the process of centralizing or aggregating data from multiple sources into one common repository.
Example Use Cases: Facebook and Retail
Data Virtualization
…is an approach to data management that allows an application to retrieve and manipulate data without requiring technical details about the data, such as how it is formatted or where it is physically located.
Example: Google
BI/Analytics
Data Warehouse (DW)
Data Warehouse Optimization with Data Virtualization
Machine Device, Cloud
Documents and Emails
Relational, Mainframe
Social Media, Web Logs
Data Sources
HDFS HDFSHDFS
Server
Deliver Enriched BI/AnalyticsVirtualize data from DW and Hadoop
Deliver richer & deeper data for Analytics
Optimize Storage / CostsMigrate infrequently used data to Hadoop
Process Unstructured, New Sources in Hadoop
Offload ELT & Virtually Expand the DW
Data Analytics and Business Intelligence
§ …is a process of inspecting, cleaning, transforming, and modeling data with the goal of discovering useful information, suggesting conclusions, and supporting decision-making.
§ Business intelligence covers data analysis that relies heavily on aggregation, focusing on business information.
Example Use Case: Retail –Inventory Accuracy
Big Data vs. Data Analytics
§ Big Data:§ Save money:
§ Cost effective scale§ Consolidate many types of data§ Operational efficiencies
§ Data Analytics:§ New ways to make money:
§ Predict, understand, and monetize customer behavior
§ Fact-based decision making§ Enable real-time tactical decisions
§ Also save money: § Increase security, reduce fraud,
predict failures
§ Business Intelligence:§ Rear-view mirror§ Answers what happened? § Tactical/departmental/siloed§ Predefined and canned§ What did we sell, to who, and
how?
§ Business Analytics:§ Windshield and
beyond§ Answers what is next?§ Strategic/enterprise/holistic§ Discovery and ad hoc§ What can we sell and to who?
Business Intelligence or Business Analytics? What’s the Difference?
12+ TBsof tweet data
every day
25+ TBs of log data every day
2.5 exabytes of data every day
30 billionRFID tags today
4.6 billioncamera phones
worldwide
100s of millions of
GPS enableddevices
sold annually
200 million smart meters
Source of the Data?
500 / Second Data sent per 4G LTE Enabled Car
How the Data Is Sourced
Where the Data Is Sourced
Remote Branch
Branch Office
Co-Location XaaS Provider
Retail Outlet
Manufacturing
• Localized Processing• “Intercloud” Connectivity• Data Optimization• Data Virtualization
Head Office
What We Need:
The Challenge
Growing Data Volumes
Shortened Processing Windows
Escalating Costs
Hitting Scalability Ceilings
Demanding Business
Requirements
ETL Complexity
Latency in Data
Tight IT Budgets
The Classic Enterprise Challenge…
…Do More with Less
Data Warehouses Cannot Cost-Effectively Support Data Growth
100% DATA GROWTH100TB 100TB 100TB
Data Warehouse$20,000 – $100,000/TB
To Add Capacity to the Data WarehouseIncremental spend of $2M – $10M
Keeps the Right Data in the Data Warehouse:• Operational Analytics• Reporting• Business Analytics
Offloading Everything Else to Big Data: Saves $1.85M – $9.8M• Historical Data• Data Processing• Data Hub/Ad Hoc Exploratory• Transformation/Batch
Today, growth is accommodated by additional investment in your data warehouse
Big Data complements your data warehouse, offloading data to defer/avoid more costly spend
100TB50TB
LOWER VALUE DATA
HIGH VALUE DATA 50TB100TB
Hadoop Cluster Cost $1000 –$2000/TBIncremental Cost $240K – $300K
Opportunity: Turning Big Data into Wisdom
Data-driven enterprises outperform their industry peers by up to 6%, are up to 26% more profitable*Gartner’s 2013 CIO survey: Analytics/business intelligence was the number one technology priority
WISDOM (Scenario Planning)
KNOWLEDGE
INFORMATION
DATA
• Collect more accurate performance data• Analyze variability, understand root causes
• Create highly specific customer segmentation• Tailor products and services• Anticipate requirements and outcomes
• Increase productivity (MIT, 2003)• Speed delivery of new innovations to market
MORE IMPORTANT
LESS IMPORTANT
*MIT, 2013
§ Big Data is a large ecosystem comprised of over ISVs. Cisco views the market as centered around three major pillars, which are data management, data warehouse optimization and expansion, and analytics and BI.
§ Data can be virtualized, warehoused, optimized, mined, analyzed, or used to produce new data.
§ CEOs, the heads of business lines, board members, marketing and financial leaders, all may be the drivers of Big Data initiatives.
Summary
Q & A