Data Warehouse Platforms A SearchDataManagement.com Directory 2009 EDITION This product brought to you by: Welcome! INTRO SELECTING THE RIGHT DATA WAREHOUSE PLATFORM FOR YOUR ORGANIZATION INDEX Welcome to SearchDataManagement.com’s Data Warehouse Platforms Directory. This directory is designed to be a valuable resource for those getting started with research or evaluating optimized data warehouses. Inside, you’ll find basic infor - 1010DATA mation about vendors and the platforms they sell. Each listing is accompanied by ASTER DATA a short description and a long description, including limited information about SYSTEMS functionality and product use. You’ll find products for businesses of all sizes as IBM well as products that can be deployed on-demand and on-premise. Use this list ILLUMINATE to get started with the evaluation process. For more information about any of the SOLUTIONS products or to speak to a sales representative, please visit the vendor website or GREENPLUM product website. INFORMATICA KALIDO SearchDataManagement.com will launch a series of directories throughout the KOGNITIO year to address unique segments of the data management market. Want to see your product listed in one of our directories? Go here to submit a product . Need to MICROSOFT CORP. update product or pricing information? Email us here . For questions for the editors NETEZZA or to make suggestions for improving the directory, write to us at editor@search - ORACLE datamanagement.com . CORPORATION PARACCEL PERVASIVE Happy shopping! SAND TECHNOLOGY SUN MICROSYSTEMS SYBASE TERADATA VERTICA ABOUT SEARCHDATA- MANAGEMENT .COM METHODOLOGY DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 2 Selecting the right data warehouse platform for your organization INTRO SELECTING THE RIGHT DATA WAREHOUSE PLATFORM FOR YOUR ORGANIZATION INDEX s companies consider I Quality data that is available. their ability to manage I Architectures that provide low 1010DATA the asset that is infor - long-term total cost of ownership. ASTER DATA I SYSTEMS mation, the data Good query performance that warehouse platform (and its database results in increased interactive IBM Amanagement system, DBMS) is the usage. ILLUMINATE I SOLUTIONS most important decision point. The The ability to get to real-time feeds. platform is the foundational component I A platform to support mixed, GREENPLUM of the tool selections, the consultancy unpredictable workloads. INFORMATICA hires, the architecture, etc. In short, it I A scalable path forward as data KALIDO defines your information culture. needs grow. KOGNITIO However, in selecting the platform to MICROSOFT CORP. support the data warehouse, organiza - tions are faced with an exponentially CRITERIA FOR DATA NETEZZA higher number of variations and distinct WAREHOUSE PLATFORM SELECTION ORACLE CORPORATION departures from the traditional online The decision process should go well transactional processing (OLTP) data - beyond the usual consideration of the PARACCEL base management systems than ever operational DBMS vendor. Today, that PERVASIVE before. decision can be nuanced along several SAND TECHNOLOGY Over time, data warehouse data vol - potential requirements, including: SUN MICROSYSTEMS umes will continue to soar as history I SYBASE accumulates, syndicated data is collect - Active loading of data and immedi - ed and new sources with more detailed ate access to loaded data. TERADATA data are added. Furthermore, the com - I A mixed workload upon the data. VERTICA munity consuming the data continues I The immediate availability of infor - to grow, expanding well beyond compa - mation. ABOUT ny boundaries to customers, supply I Cross-functional complexity. SEARCHDATA- I MANAGEMENT chain partners and even the Internet. The level of query concurrency. .COM Companies need to make sure they I The scalability needs of the plat - METHODOLOGY choose a proven platform not just for form. the initial, known requirements but also I The functionality of the DBMS. with the ability to scale to future, to-be- determined requirements. Given the state of the marketplace, Data warehouses, to be successful, the technical architecture for a data need to provide: warehouse platform should be: DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 3 Scalable: The data warehouse plat - Available: The platform should sup - form should be scalable in both per - port mission-critical business applica - formance capacity and incremental tions with minimal downtime. Check on INTRO data volume growth. Make sure the pro - “hot pluggable” components, understand posed platform scales in a near-linear system downtime requirements and SELECTING THE RIGHT DATA fashion and behaves consistently with any issues that might deny or degrade WAREHOUSE PLATFORM growth in all of database size, number service to end users. These can include FOR YOUR of concurrent users, and complexity of batch load times, software/ hardware ORGANIZATION queries. Understand the additional upgrades, severe system performance INDEX hardware and software required for issues, and system maintenance outages. each of the incremental uses. 1010DATA Interoperable: The platform should ASTER DATA Powerful: The data warehouse plat - allow for integrated access to the Web, SYSTEMS form should be designed for complex internal networks and corporate main - IBM decision support activity in a multi-user frames. ILLUMINATE mixed workload environment. Check on SOLUTIONS the maturity of the optimizer for sup - Affordable: The proposed technology GREENPLUM porting every type of query with good (hardware, software, services, required INFORMATICA performance and to determine the best customer support) should allow for low execution plan based on changing data total cost of ownership (TCO) over a KALIDO demographics. Check on conditional multi-year period. KOGNITIO parallelism and what the causes are of MICROSOFT CORP. variations in the parallelism deployed. Flexible: The platform should be able NETEZZA Check on dynamic and controllable pri - to provide optimal performance across ORACLE oritization of resources for queries. the full range of normalized, star and CORPORATION hybrid data schemas with large num - PARACCEL Manageable: The platform should be bers of tables. Look for a proven ability PERVASIVE manageable with minimal support tasks to support multiple applications from requiring DBA/System Administrator different business units, leveraging data SAND TECHNOLOGY intervention. It should provide a single that is integrated across business func - SUN MICROSYSTEMS point of control to simplify system tions and subject areas. SYBASE administration. You should be able to TERADATA create and implement new tables and Robust in-database management VERTICA indexes at will. systems, features and functions: The platform should include DBA productiv - Extensible: The data warehouse plat - ity tools, monitoring features, parallel ABOUT SEARCHDATA- form should provide flexible database utilities, robust query optimizer, locking MANAGEMENT design and system architecture that schemes, security methodology, intra- .COM keeps pace with evolving business re- query parallel implementation for all METHODOLOGY quirements and leverages existing in- possible access paths, chargeback and vestment in hardware and applications. accounting features, and remote main - Ask such questions as: What is required tenance capabilities. to add and delete columns? What is the impact of repartitioning tables? Referenceable: There may not be one DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 4 single reference that matches your ORIENTATION environment exactly, but you should be A row-oriented database is an imple - able to see a consistent trend across a mentation of relational theory. Data INTRO wide range of references that enforces is stored as bytes, with all columns for what you are looking for in a data ware - a row stored in order. These bytes are SELECTING THE RIGHT DATA house platform. grouped by the several thousand—from WAREHOUSE PLATFORM Organizations should also consider 4,000 to 64,000—into data blocks. FOR YOUR vendor viability, especially in these This is the unit of input/output, an ORGANIZATION days of vendor consolidation. The ven - exception being the “pre-fetch” capa- INDEX dor’s financial stability, the importance of data warehousing to their overall 1010DATA business strategy and their continued “Organizations should ASTER DATA research and development in these SYSTEMS areas toward a well-developed and also consider vendor IBM relevant vision are all key components viability, especially in ILLUMINATE of a vendor’s viability in this critical SOLUTIONS decision. these days of vendor GREENPLUM consolidation.” INFORMATICA DATA WAREHOUSE —WILLIAM MCKNIGHT KALIDO ARCHITECTURAL CONSIDERATIONS KOGNITIO To enable the above criteria, companies MICROSOFT CORP. have some new options in the types of bility of row-oriented database man - NETEZZA data warehouses available, many of agement systems (DBMS) sensing a ORACLE which have been created in recent years pattern in the reads. A columnar DBMS CORPORATION owing to the high cost of ownership of is an implementation of the relational PARACCEL the traditional approaches. theory, but with a twist. The data stor - PERVASIVE Today, typical data management age layer contains a grouping of environments have several restrictions columns. For example, all the column 1 SAND TECHNOLOGY they’ve "learned to live with” in seeking values are physically together, followed SUN MICROSYSTEMS acceptable analytical performance. by all the column