IBM PureData System for Hadoop Partner Program Jan Shineman, Director, Alliance Management
by user
Comments
Transcript
IBM PureData System for Hadoop Partner Program Jan Shineman, Director, Alliance Management
IBM PureData System for Hadoop Partner Program Jan Shineman, Director, Alliance Management Brian Hess, Director, Advanced Analytics 1 © 2013 IBM Corporation If this were easy, everyone would already be leveraging big data “Big Data offers big business gains but hidden costs and complexity present barriers that most organizations will struggle with” - The Cost of Big Data, Eric Savitz, Forbes 5/2012 Open source Apache Hadoop implementation for enterprise usage is incomplete Hadoop skills are in short supply Custom built solutions lack integrated cluster management Requires integration effort within the existing analytic ecosystem Most integrated solutions do not help with archival 2 © 2013 IBM Corporation Table 1 Top Main Worldwide Hadoop-MapReduce Ecosystem Software Revenue, 2010–2016 Revenue ($M) Growth (%) 2010 38.5 2011 77 2012 209.2 2013 378 2014 539 2015 682.5 NA 100 171.7 80.7 42.6 26.6 2011–2016 2016 CAGR (%) 812.8 60.2 19.1 Note: See Table 2 for top 3 assumptions and Table 3 for key forecast assumptions. Source: IDC, 2012 Figure 2 Worldwide Hadoop-MapReduce Ecosystem Software Revenue, 2010–2016 Top Main “Currently, Hadoop in general and MapReduce in particular require special skills and a great deal of manual effort. This means a great risk of error and a shortage of people to do the job. Over the forecast period, tools will be made available to build and manage these solutions, thus amplifying demand for Hadoop-MapReduce software overall.” IDC 2012 3 © 2013 IBM Corporation Agenda IBM PureData for Hadoop – Overview and Launch Technology and Partner Integration Next Steps Q & A 4 © 2013 IBM Corporation Announcing the new PureData System for Hadoop Accelerate time to value System for Hadoop Accelerate time to insight Simplify big data adoption and consumption Extend the value of the data warehouse Implement enterprise class big data Minimize system setup and administration 5 © 2013 IBM Corporation IBM PureData System for Hadoop Accelerate Hadoop analytics with appliance simplicity System for Hadoop Speed Speed to insight with built-in analytics Speed to value with accelerated deployment Simplicity Ready to load data in hours Integrated system management Appliance approach reduces complexity Single point of support Smart Establish a cost efficient online data archive Easily leverage data across the big data platform Enterprise security, governance and high availability 6 © 2013 IBM Corporation PureData System for Hadoop Enables Key Use Cases Enrich Your Information Base Improve Customer Interaction with with Big Data Exploration Enhanced 360º View of the Customer 99% Reduction In Time Required For Analysis 1,100 with Operations Analysis 600K Metered Customers in Five States 7 with Security and Intelligence Extension 42TB Association Publishing Partnerships Optimize Infrastructure and Monetize Data Prevent Crime Real-time Acoustic Data Analyzed Gain IT efficiency and scale with Data Warehouse Augmentation 40X Gain in Analysis Performance © 2013 IBM Corporation Use Case: Big Data Exploration Use Cases • Explore new data and previously untapped sources • Visualize and gain new insight with easy to use spreadsheet-style analysis • Identify useful information that would add value when integrated • Used for data profiling to understand data before moving to other systems 8 © 2013 IBM Corporation Use Case: DWA Pre-Processing Hub Use Cases • Aggregation of data • Pre-process cleansing • Compliance requirements • Simple analytics / exploration 9 © 2013 IBM Corporation Use Case: DWA Active Archive PureData System for Hadoop PureData System for Analytics Use Cases • Immediate storage alternative of cold data • Cost savings for cold data • Compliance requirements • Simple analytics / exploration 10 © 2013 IBM Corporation BDIG: Launch Event Agenda Start End Title 8:30 9:00 Registration & Breakfast 9:00 9:10 Welcome Eric Sall, VP IM Marketing 9:10 9:30 Building Confidence in Big Data Bob Picciano, GM IM 9:30 10:00 Leverage Confidence within Your Big Data Use Case Inhi Cho Suh, VP, Big Data, Integration & Governance 10:00 10:30 Applying Integration and Governance to Big Data Use Cases Michele Goetz (Forrester) 10:30 10:45 Break 10:45 11:20 Innovations in IIG: Automation, Context, and Agility Martin Wildberger, VP IM Development 11:20 11:35 Demonstration: Building Confidence in Big Data Guenter Sauter, IM Product Management 11:35 12:10 Client Panel – How to Build Confidence in Big Data Inhi Cho Suh 12:10 12:15 Wrap up Eric Sall 12:15 1:15 Lunch 1:15 3:00 1:1s with thought leaders &IBM Execs Event Description September 10th, 2013 590 Madison Executive Briefing Center, New York Launch event type: Exclusive, by-invitation only, executive level event Local Analysts & Influencers Target # of attendees 30-40 Building Confidence in Big Data Confidence in your Data InfoSphere builds confidence with trusted & protected data • • • Automated Integration with Data Click Visual Context via the Governance Dashboard Agile governance to protect big data with InfoSphere Guardium Confidence in Accelerating Value PureData for Hadoop simplifies Hadoop. • • Appliance simplicity for Hadoop systems, Get up and running in hours Confidence in your Skills Stampede delivers everything needed to start with Big Data • • 11 All the resources needed to get value from big data quickly Software, Expertise, Training Speaker © 2013 IBM Corporation From Getting Starting to Enterprise Deployment: InfoSphere BigInsights Brings Hadoop to the Enterprise PureData for Hadoop Enterprise class - Appliance simplicity for the enterprise BigInsights Enterprise Edition Sold by # of terabytes managed BigInsights for Hadoop QuickStart Edition Free download Apache Hadoop 12 - Web-based mgmt console - Jaql - Integrated install - Many enterprise features - Tutorials - Accelerators - Performance Optimization - Visualization Capabilities - Pre-built applications - Text analytics - Spreadsheet-style tool - RDBMS, warehouse connectivity - Administrative tools, security - Eclipse development tools - Enterprise Integration .... Breadth of capabilities © 2013 IBM Corporation IBM BigInsights Leads the way for Hadoop Platforms Big Data Platforms look at data end to end – no just in Hadoop “IBM has the deepest Hadoop platform and application portfolio. IBM, an established EDW vendor, has its own Hadoop distribution; an extensive professional services force working on Hadoop projects; extensive R&D programs developing Hadoop technologies; connections to Hadoop from its EDW.” –The Forrester Wave™: Enterprise Hadoop Solutions, 1Q12 13 © 2013 IBM Corporation Agenda IBM PureData for Hadoop – Overview and Launch Technology and Partner Integration Next Steps Q & A 14 © 2013 IBM Corporation Offering Highlights for IPDH V1 Current issues customer face… … and what we plan to deliver IBM InfoSphere BigInsights High Data Growth • • • • • Unreliable Hadoop – subprojects suite, Jaql Analytics – built-in text analytics & tooling Usability – web console management Enterprise Class – security, cluster mgmt. Integration – DB2, Netezza, JDBC DBs Netezza Appliantization Support Complexity • Manageability – appliance, cluster mgmt. • Appliance level serviceability – replacement, update, support Lack of Analytics IBM System X Hardware High Support Costs Silo System • Reliable, enterpriseready appliance • Easy-to-use, manageable system • • • • Management Node – x3550 M4 Data Node – x3630 M4 Rack Switch – BNT 8264 Management Switch – BNT 8052 • Built-in analytics 15 15 © 2013 IBM Corporation Software BigInsights Enterprise Edition –Based on version 2.1 • Hadoop v1.0, Zookeeper, Hive, HBase, Oozie, etc • BigSQL, Console, JAQL, BigSheets, Accelerators, etc –Update Plan: Incorporate BigInsights releases within a quarter System Management –New “Hardware” tab to manage hardware –Component level view (server, disk, fan, CPU, etc) High Availability –Protects critical Hadoop services (NameNode, JobTracker, etc) –HA Master node via Linux HA EasyArchive+ –New “EasyArchive+” tab and archive/retrieve Console apps –Extract data from Netezza to HDFS and surface via Hive tables –Return data from HDFS archive back to Netezza 16 © 2013 IBM Corporation IBM Open Source BigInsights Enterprise Edition Components IBM InfoSphere BigInsights Development Tools Visualization & Discovery Eclipse Plug-ins BigSheets Text Analytics MapReduce Jaql Dev’t Hive Query Systems Management Web Admin Console Connectors JDBC Netezza Advanced Engines Text Processing Engine and Extractor Library Workload Optimization Integrated Installer Splittable Text Compression ZooKeeper Oozie Jaql Lucene Pig Hive Data Store File System DB2 Streams Enhanced Security Runtime System ML Adaptive MapReduce Flexible Scheduler R Flume BigIndex Sqoop MapReduce HBase Column Store HDFS GPFS (Beta) © 2013 IBM Corporation BigInsights Value Above and Beyond Hadoop Analytic Accelerators – Social Media Accelerator – Machine Data Accelerator – BigSheets spreadsheet and visualization – Advanced Text Analytics Accelerator – JAQL query language Performance and Optimization – Adaptive Map Reduce – Advanced Scheduler – BigIndex for large scale indexing – Fast, splittable compression Security – Role based authorization Optim Development Studio – Eclipse based IDE for Java Big Data Integration – Information Server, InfoSphere Streams, Netezza, DB2 Enterprise Enablement – Big SQL – GPFS-FCO IBM’s distribution is based on Apache Hadoop and utilizes many of the capabilities includes in that distribution, but IBM is focused on making its distribution more of an enterprise class offering.” 18 © 2013 IBM Corporation IPDH Architectural Highlights Master nodes Data nodes Appliance Management HA Appliance hardware management High-Availability (via failover) on master nodes BigInsights Unified GUI Complete self-contained BigInsights appliance Redundant components for failure resilience (disks, switches, power supplies) Secure operation (access only via edge nodes) GUI, CLI, and API for administrative controls 19 19 © 2013 IBM Corporation Hardware Overview Full Rack Data Nodes Memory per node 96 GB CPU Cores per node 12 Drives per node 12 Drive Size 20 18 3 TB Total Raw Storage 648 TB User Space (w/ replication) 216 TB © 2013 IBM Corporation Hardware Specifications GTM Master Node (x2) Data Node (x18) Rack Switch x3550 M4 •Dual Intel E5 2600 series •128 GB RAM •3 x 3 TB 3.5” Drives •2 x 10Gbe ports •4 x 40Gbe ports x3630 M4 •Dual Intel E5 2400 series •96 GB RAM •14 x 3 TB 3.5” Drives •Dual Port 10 Gbe BNT 8264TR (x2) •48 x 10 Gbe •4 x 40 Gbe Management Switch BNT 8052 (x1) •48 x 1 Gbe Designed for High Availability © 2013 IBM Corporation Application Integration (IPDH version 1.0) Secured configuration – Default – Access to the system via BigInsights REST APIs – ODBC/JDBC for BigSQL and Hive – REST API for: • HDFS read/write • Console app execution – Access to Master Node services: • Oozie, Zookeper, Hive, etc – No access to Data Nodes, so: • No HDFS RPC API (or libhdfs) • No HBase RegionServer access Connected Client configuration – Access to Data Nodes and Master Node • HDFS NameNode, JobTracker, Zookeper, Hive, HBase Master, etc • AND HDFS RPC API, HBase RegionServer, etc. – Should support all Hadoop applications 22 © 2013 IBM Corporation Agenda IBM PureData for Hadoop – Overview and Launch Technology and Partner Integration Next Steps Q & A 23 © 2013 IBM Corporation IPDA Technical Validation Process and Acceptance into “Ready For PureSystems” Program Registration Obtain IBM ID and Register your Company in IBM PartnerWorld Execute IPDA /IPDH Attachment in IBM PartnerWorld Technical Validation Request Access to Hosted Partner Network or Download Big Insights 2.1 Use your usual test methodology for verifying data source capability “Ready for” PureSystems IBM will confirm your acceptance into Ready For Pure Systems Program Sign the online supplement that grants authorization to use the Ready For mark Contact us with questions Enter your solution information into Global Solution Directory 24 Confirm validation of specific versions to technical management Use the mark on your web site and other marketing efforts. © 2013 IBM Corporation Next Steps IPDH is accessible in the partner lab – [email protected] No emulator at this time Big Insights 2.1 is downloadable (see next slides) Completion qualifies your solution for the PureSystems Center (ReadyFor IPDH) 25 © 2013 IBM Corporation Value Package - PartnerWorld Software Benefits Purchase Required IBM Value Package $2K US subscription $1.8K US renewal – Software access – Pre-sales pre-deployment technical support – You Pass, We Pay • Designed for Business Partners invested in IBM software • Critical for attaining/retaining Software Value Plus authorization • Tiered by membership level Software Access Option – Software 26 © 2013 IBM Corporation Software Access Option Annual subscription $795 US – Access to IBM Software; usage includes • Demonstration & evaluation • Development and testing • Internal training • Run Your Business www.ibm.com/partnerworld/value 27 © 2013 IBM Corporation Summary Value for partners: The Hadoop market is forecast to experience a CAGR of over 60% by 2016. (IDC, 2012) The IBM PureData System for Hadoop brings appliance simplicity to customers and partners for enterprise class Hadoop deployments. IBM’s focus and leadership on Big Data can help business partners grow sales and expand target markets. 28 © 2013 IBM Corporation Questions? 29 © 2013 IBM Corporation © International Business Machines Corporation 2012 International Business Machines Corporation New Orchard Road Armonk, NY 10504 IBM, the IBM logo, PureSystems, PureFlex, PureApplication, PureData and ibm.com are trademarks of International Business Machines Corporation, registered in many jurisdictions worldwide. A current list of IBM trademarks is available on the Web at www.ibm.com/legal/copytrade.shtml All rights reserved. WAP12402-USEN© 2013 IBM Corporation 30 01