SAS and Oracle: Big Data and Cloud Partnering Innovation Targets the Third Platform David Lawler, Oracle Senior Vice President, Product Management and Strategy Paul Kent, SAS Vice President, Big Data
What is the 3 rd Platform? IDC: Platform built on Mobile, Social Networking, Cloud, Big Data Analytics Data driven society: Data is Value Data is now a product, a service, a medium of exchange Ability to analyze and act upon data creates competitive advantage Big Data & Analytics - Top Investment Priority for organizations of all sizes Source: Dan Vesset, IDC 2
SAS & Oracle: Leading the 3 rd Platform SAS Leadership in Data Management Analysts recognize SAS leadership Advanced Analytics Platform Data Integration Data Quality Enterprise Business Intelligence Platforms Big Data Predictive Analytics Solutions Source: Gartner Magic Quadrant, Forrester Wave Reports 3
SAS & Oracle: Leading the 3 rd Platform Oracle Leadership in Data Management 9 of the Top 10 Cloud Companies run on Oracle Strongest Infrastructure Business in the Industry Increasing Market share: Oracle +10% in a -5% market Complete portfolio for the Enterprise Differentiated value: focused on data Engineered systems growing 30% SuperCluster growing by triple digits for 3 rd quarter in a row 4
5
6
Data Drives Business Your Business Open 7
Oracle Started as a Database Company Focused on the data that drives database success 8
Data Integrity, Reliability, Performance Oracle Engineering Values: Maximize customer value 9
Oracle Adapted to the Systems Landscape Software innovation for overall system value 2000: Hardware vendors charge high-end premium Cost often over 10x higher per socket on high end systems Limits Oracle revenues and growth Oracle re-engineers our database: Oracle RAC Option to use less expensive hardware scaled horizontally Significantly lower cost per database transaction Common data store in shared network storage Drive the highest value for customers 10
Transform System Boundaries Engineered Systems 2008 Exadata: Redefine System Distributed Storage Distributed storage across the network Removed storage bottleneck Application attached storage Integrated System Increase reliability and reduced time to production Dramatically increased performance and value by rethinking the system 11
The Acquisition of Sun Closed on January 27 th, 2010 12
Oracle Engineered Systems for ExaData ExaLogic SuperCluster Big Data Appliance ZFS Storage Appliance Virtual Compute Appliance Database Backup, Recovery, Logging Appliance 13
Thousands of Engineered Systems Customers Existing Applications, Spectacular Results 14
Oracle Engineered Systems Taking In-Memory Computing to a New Level Data Set Sizes Are Exploding Users Want Real Time Analytics Changing Memory Economics 15
Oracle Database 12c Breakthrough In-Memory Database Technology BOTH row and column in-memory formats for same data/table Memory Memory Simultaneously active and transactionally consistent OLTP Sales Row Format Sales Column Format Analytics 100X Faster Analytics & reporting: column format 2X Faster OLTP: row format 16
While testing Oracle Database In-Memory I measured query speeds 100X faster than the same queries in-memory today." Sudhi Vijayakumar Senior Oracle DBA Yahoo Inc. 17
Unprecedented Performance and Scalability Oracle M6: Terabyte Scale Computing 3 TERABYTES PER SECOND SYSTEM BANDWIDTH 1.4 TERABYTES PER SECOND MEMORY BANDWIDTH 1 TERABYTE PER SECOND I/O BANDWIDTH 384 cores, 3,072 threads Hardware Optimized Virtualization 32 Terabytes Memory 18
Oracle: Historical Re-engineering Economics of the SMP Economics of SMP $/Unit of Performance Near Linear Pricing IBM Power 740 POWER7+ $101,571 IBM Power 750 POWER7+ $204,982 IBM Power 780 POWER7+ $2,101,370 IBM Power 795 POWER7 $6,491,183 Worse Better T5-2 $67,042 T5-4 $147,992 T5-8 $268,314 M6-32 $1,209,943 2 Socket 4 Socket 8 Socket 32 Socket 19
Dramatic Scalability for Analytics "SPARC and Solaris have long been a proven platform for SAS applications. We've seen in-house that the technically-advanced features and design of the SPARC M5 servers along with processor and throughput enhancements, provide a very wellsuited platform for enterpriseclass SAS application deployments. Craig Rubendall Senior Director of R&D SAS M5 5,000 SAS users Consolidated 7 systems to 2 30% performance improvement 70% reduction in processors Lower licensing costs Plan to consolidate 50+ Oracle DBs to same system 20
Oracle M6-32 SuperCluster In-Memory Database & Application System Fastest Database Machine Big Memory for Column Store 32 Terabytes of Memory 3 Terabyte Silicon Network Integrated Exadata Storage for 10x Database I/O Acceleration InfiniBand I/O Interconnect 21
Investing in Silicon Leadership ACQUIRE x86 IBM 2010 SPARC SPARC T-Series FOCUS Oracle 100% performance each generation 2011 SPARC COMPETE 2012 ACCELERATE SPARC IBM Power & x86 30 50% performance each generation 2013 OPTIMIZE X86 & IBM: incremental improvements Next Generation Oracle SPARC Oracle Processors doubling performance every 2 years FUTURE 22
Investing in Silicon Leadership ACQUIRE FOCUS COMPETE ACCELERATE OPTIMIZE Oracle 100% performance each generation SPARC M7/T7 M8/T8 SPARC M7/T7 Running in lab now x86 IBM x86 IBM SPARC SPARC SPARC T-Series SPARC SPARC IBM Power & x86 30 50% performance each generation Deep Software in Silicon Step function in performance M8/T8 New core 2010 2011 2012 2013 FUTURE 23
The Ultimate Software Optimization: Hardware Moving Oracle Database & Java Software Functions into Hardware Software in Silicon Data query acceleration Java acceleration Application data protection Data decompression Only Oracle 100 s running In labs now 24
The 3 rd Platform in Business Big Data Data Base Analytics 25
Unstructured Data Sources Big Data Unstructered data Multiple sources Identify high-value relationships Define structured extracts Data Base Analytics 26
Real Time Analytics Big Data Draw from structured & unstructured data Identify core business value Real-time performance optimized In memory Data Base Analytics 27
Apply Real-Time Business Rules Structured inflows Core business data Transactional & Analytic Performance optimized In memory Big Data Data Base Analytics 28
One Size Does NOT Fit All IDC: No Two Companies are Exactly Alike Not just Hadoop, Social Media data, Data Scientists Different: Workloads, data types, user types, corporate environments Look carefully at your environment Use the technology purpose built for specific use cases 29
Oracle Engineered Systems Simplify Your Environment Extreme Performance Low Risk Deployment Extreme Efficiency 30
SAS OnDemand Runs on Exadata SAS IT Runs Exadata Business Continuity: High Availability SLA s >99% Superior Backup, restore, and recovery High value: Decreased time to production for new customers No need to take down to expand capacity Best of Breed: Management, monitoring, development and diagnosis Resource Management/Optimization Security(HW to SW) capabilities 31
SAS & Oracle: Working Together Building Our Strongest Joint Partnership Ever Extensive engineering collaboration Sizing, configuration guidance and best practices for deployment Joint support for Proof of Values/Proof of Concepts Strong technology and business alliance Develop solutions and products brings tremendous value and confidence SAS High-Performance Analytics and SAS Visual Analytics on Oracle Engineered Systems 32
SAS and Oracle: Big Data and Cloud Partnering Innovation Targets the Third Platform Paul Kent, SAS Vice President, Big Data David Lawler, Oracle Senior Vice President, Product Management and Strategy
Agenda Best of Breed Partnering Big Data @ SAS SAS & Oracle My Big Data Challenge 34
Best of Breed Partnership Innovation 35
Reflection on a stronger partnership than ever SAS High-Performance Analytics and SAS Visual Analytics on Oracle Engineered Systems Extensive engineering collaboration Sizing, configuration guidance and best practices for deployment Support for POVs A strong technology and business alliance that brings tremendous value and confidence to customers 36
Oracle Big Data Appliance Buy vs Build Advantages Initial Cost and Time to Value (ESG Analyst: 40% cost savings, 8 weeks faster implementation ) Performance and Scalability Support and maintenance effort Pre-configured with leading Hadoop Distribution Starter Rack In-Rack Expansion Sun Oracle X3-2L Servers Per node: Full Rack 2 sockets, 16 cores Intel Xeon 512 GB Memory Proven at large scale Contributors across all components for better support Better Integration with your Oracle ecosystem High-performance connectivity to Exadata Single Enterprise Manager Framework 12 x 3 TB Disks 37
What is a Data Scientist? 38
Josh Wills 39
OUR PERSPECTIVE Big Data is RELATIVE not ABSOLUTE BIG DATA ANALYTICS BIG ANALYTICS When volume, velocity and variety of data exceeds an organization s storage or compute capacity for accurate and timely decision-making The process surrounding the development, interpretation, and useful application of statistics to solve a problem. Analytics applied to data provides the 4 th V = Value Three types: Descriptive, Predictive, Prescriptive The combination of using ANALYTICS on BIG DATA AND/ OR the capability to run advanced or complex analytics on any size data. 40
SAS Analytics Lifecycle 41
SAS on Hadoop 42
Today 43
Big Picture 44
Tomorrow 45
SAS BUSINESS ANALYTICS FRAMEWORK SAS Scoring Accelerator on Oracle Database and Exadata Available since June 2012 EXADATA Enterprise Data Warehouse with SAS In-Database Analytics + + Relational Data Store Analytic Data Warehouse / Marts Primary Failover SAS Analyst s Desktops SAS Compute Nodes Web Application Server(s) SAS Web Clients Data Tier Server Tier Metadata Tier Web Tier Client Tier Copyright 2012, SAS Institute Inc. All rights reserved. 46
SAS BUSINESS ANALYTICS FRAMEWORK SAS Visual Analytics, SAS High-Performance Analytics Server for Oracle Engineered Systems Available since Dec 2012 EXADATA Enterprise Data Warehouse with SAS In-Database Analytics + + Relational Data Store Analytic Data Warehouse / Marts Data Tier Oracle Big Data Appliance Exalogic, Virtual Compute Appliance SAS Compute Nodes Primary Failover Web Application Server(s) SAS Analyst s Desktops SAS Web Clients Server Tier Web Tier Client Tier Metadata Tier Copyright 2012, SAS Institute Inc. All rights reserved. 47
SAS High Performance Analytics, SAS Visual Analytics on Oracle Engineered Systems Big Data Appliance (BDA) Exadata SAS Analyst s Desktops bda101 bda102 SAS High-Performance Analytics Server Root Node SAS Visual Analytics Server Tier Hadoop Namenode Hadoop Datanode bda103- bda118 SAS Visual Analytics Middle Tier SAS LASR In-Memory Analytics Server SAS Web Clients 48
SAS High-Performance Analytics - Choice Distributed Scale Out Big Data Appliance, Exalogic, OVCA Oracle Linux SMP Scale Up SPARC M5-32, Solaris 11 SAS 9.4 on Solaris LASR In-Memory Analytic Server High Performance Analytics Server High Performance procedures (ie: hplogistic, hpreg, hpreduce, hpdmdb.) Infiniband 49
My Big Data Challenge to You SAS & Oracle on Big Data Appliance (last year) SAS In-Memory Statistics (IMSTAT) Announcement Challenge to you Production bringup with dedicated resource support, up here at the Analytics Conference or SAS Global Forum 2015 with me and Oracle 50
Paul Kent, SAS VP Big Data paul.kent@sas.com @hornpolish