WIXLIB

IBM PureData SystemIBM PureData Systemfor Operational AnalyticsAn integrated, high-performance data system foroperational analyticsHighlights Provides an integrated, optimized,ready-to-use system with built-inexpertise for operational analytics Delivers outstanding performance andthroughput for in-database analysis oflarge data sets that include both historicand operational data Continuously ingests data to supportnear-real-time responsiveness to dynamicbusiness environments Designed as a modular, scalable systemthat can grow with your business Designed to handle more than 1,000concurrent operational queries1 Integrated and simplified monitoringand maintenance Compatible with market-leadinganalytic and BI tools, applications andinfrastructure Powered by IBM DB2–basedIBM InfoSphere Warehouse softwareand IBM POWER7 processor-basedIBM Power Systems serversBusinesses across industries need actionable insight into theiroperations to gain a competitive advantage. To gain this insight,they often must analyze large data sets comprising both historicand current (or operational) data. However, a high percentage ofqueries for operational analytics systems—often up to 80 percent—are interactive lookups focused on data about a specific customer,account or patient. These operational queries can originate from callcenters, mobile sales apps, real-time fraud detection systems or otherapplications that support real-time decision making.To deliver the correct information as rapidly as possible, the datawarehouse supporting these systems must be optimized for the rightbalance of analytics performance and operational query throughput.The IBM PureData System for Operational Analytics—a memberof the IBM PureSystems family—helps organizations meet thesecomplex requirements with an expert integrated data system that isdesigned and optimized specifically for the demands of an operationalanalytics workload. Built on IBM Power Systems servers with IBMSystem Storage and powered by IBM DB2 –based InfoSphere Warehouse software, the system is a complete, out-of-the-box solutionfor operational analytics that provides both the simplicity of anappliance and the flexibility of a custom solution. Designed to handlemore than 1,000 concurrent operational queries,2 it delivers missioncritical reliability and scalability with outstanding performance.

IBM PureData SystemTurn data into insight with in-databaseanalyticsWith in-database analytics, you run analytics on your datawhere it resides—in the warehouse. This eliminates thetime, cost and risk associated with copying data out of thewarehouse to analyze it.Using multidimensional cubing services, this PureData Systemdelivers rapid insight into high volumes of fast-moving data.Users can create, edit, import, export and deploy cube modelsover the relational warehouse schema to analyze multiplebusiness variables. Cubing services help optimize performancefor online analytical processing (OLAP) queries, providingmore power for users to analyze data and generate businessinsight that can enhance both profitability and customersatisfaction. Overall, performance for complex queries is upto 3.3 times faster than the previous release of InfoSphereWarehouse software.3Powerful data mining capabilities also enable integratedanalytics of both structured and unstructured data in thesystem. Standard data mining models (including clustering,associations, classification and prediction) are supported andcan be developed via drag and drop in an intuitive designenvironment. The data mining models are executed in theproduction environment to provide real-time scoring ofdata records. Additionally, rich presentation componentsare provided to enable visual analysis of data mining results.The IBM PureData System for Operational AnalyticsGet real-time analytical insights withcontinuous data ingestThe PureData System for Operational Analytics enablesIT departments to easily deploy, optimize and managedata-intensive workloads for operational analytics. Itdelivers exceptional value in three ways: Continuous data ingest capabilities enable organizationsto transparently load data from external sources into thePureData System without downtime—supporting real-timebusiness analysis and decision making during the loadingprocess. The continuous data ingest feature allows ITdepartments to load data across multiple threads at the sametime to get the data in very quickly, while also dynamicallyswitching between the various external load sources to helpmaximize resource utilization. It helps eliminate the latencycreated by batch-loading data on infrequent schedules,making it extremely valuable for business users who needcurrent operational data in the warehouse.Built-in operational analytics expertise, based on years ofIBM experience and best practices from thousands of clientengagements, is embedded into the system to providea complete solution that can rapidly deliver value.Integration by design of software, server, storage andnetworking results in factory-optimized systems designedfor fast time-to-value, efficiencies and high performance.A simplified experience from design to purchase tomaintenance helps reduce total cost of operations.2

IBM PureData SystemGrow as needed with a flexible andefficient system designKey capabilitiesA modular, flexible system design enables organizations toacquire the PureData System for Operational Analytics at thecorrect size for their current needs and scale incrementally upto a petabyte of capacity as their data grows.4Database managementIBM DB2 9.7 or 10 Enterprise EditionContinuous data ingestThe system includes many features that are preset tooptimize performance, throughput and resource utilizationof operational analytics workloads. It is also designed tosignificantly reduce disk space requirements and improvequery performance. Using Adaptive Compression features, thePureData System for Operational Analytics can automaticallycompress indexes and temporary tables to help reduce storagecosts. Data row compression contributes to storage spacesavings and helps reduce I/O overhead—and the stored pagesare also compressed, which further enhances the compressionon disk. Because data is compressed, it significantly reducesthe I/O requirements and helps improve query response timewithout the need to frequently reorganize the data. AdaptiveCompression can also adapt to changing patterns in the data.Clients have experienced cases of 10x storage space savings viaAdaptive Compression.5Storage compression with Adaptive CompressionRely on highly available, highperformance operational analyticsIBM Cognos Business Intelligence (5 user entitlements)Label- and Row-Based Access ControlIBM Workload ManagerData movementIBM SQL Warehousing ToolIBM InfoSphere Federation ServerOperating systemIBM AIX 7.1AnalyticsCubing servicesText analyticsThe fault-tolerant design of the PureData System forOperational Analytics virtually eliminates single pointsof failure and includes standby server capacity to supportcontinued operations in the event of a hardware failure.With built-in automated workload management features,IT departments can establish and enforce service levels forend users by prioritizing queries from different users andapplications and then controlling the number of underlyingresources dedicated to those processes.Intelligent minerToolingIBM PureData System ConsoleIBM Design StudioIBM Optim Development Studio3

IBM PureData SystemTable 1: IBM PureData for Operational Analytics configurationsExtra SmallSmallFoundation rack 1/3 rack1 foundation node1 data nodeMediumLargeFoundation rack1 foundation moduleCores32648096Memory256 GB512 GB640 GB768 GBSSD storage4.8 TB9.6 TB12 TB14.4 TBHDD unformatted raw capacity64.8 TB151.2 TB237.6 TB324 TBHDD RAID capacity54 TB126 TB198 TB270 TBUser data capacity - uncompressed29.7 TB69.3 TB108.9 TB148.5 TBPrimary servers1234Standby servers1222Specification: 3.2 GB/s for foundationnode, 6.4 GB/s for data node3.2 GB/s9.6 GB/s16 GB/s22.4 GB/sDatabase disk IOPS34 KB57 KB148 KB205 KBData load rateUncompressed: 1,161 GB/h Uncompressed: 3,484 GB/h Uncompressed: 5,807 GB/h Uncompressed: 8,130 GB/hCompressed: 890 GB/hCompressed: 2,670 GB/hCompressed: 4,450 GB/hCompressed: 6,230 GB/hFoundation 2/3 rack1 foundation node2 data nodesFoundation full rack1 foundation node3 data nodesHDD storageUncompressed disk bandwidthDatabase software and toolsIBM data warehousing and analytics software entitlements includedProcessors and operating systemIBM POWER7 with AIXPower (watts maximum)Foundation rack: 6,196Foundation rack: 6,196Data rack: 4,647Foundation rack: 6,196Data rack: 7,551Foundation rack: 6,196Data rack: 10,454Typical cooling (BTU/hour)Foundation rack: 14,160Foundation rack: 14,160Data rack: 11,601Foundation rack: 14,160Data rack: 19,534Foundation rack: 14,160Data rack: 27,467WeightFoundation rack: 1,450 lbs(658.3 kg)Foundation rack: 1,450 lbs(658 kg); data rack:1,250 lbs (567.5 kg)Foundation rack: 1,450 lbs Foundation rack: 1,450 lbs(658 kg); data rack:(658 kg); data rack:1,650 lbs (749.1 kg)2,150 lbs (976.1 kg)Rack dimensions (W x D x H)644 mm (25.4 in) x 1,465 mm (57.7 in) x 2,015 mm (79.3 in), including doorsVoltage drops/rackDrops/rack200-240 V ac; frequency: 47-63 Hz4 x 30A4 x 30A and 4 x 60ASafetyEmissions4 x 30A and 4 x60A4 x 30A and 4 x 60AIEC 60950-1; UL 60950-1; CSA 60950-1CISPR 22; CISPR 24; FCC, CFR 47, Part 15 (US); VCCI (Japan); Directive 2004/108/EC (EEA); ICES-003, Issue 4 (Canada); ACMA radio communicationsstandard (Australia, New Zealand); CNS 13438 (Taiwan); Radio Waves Act, MIC Rule No. 210 (Korea); Commodity Inspection Law (China); TCVN 7189(Vietnam); MoCI (Saudi Arabia); SI 961 (Israel); GOST R 51318.22, 51318.24 (Russia).4