Efficiently Maximizing Retail Value Across Distributed Data
Transcription
Efficiently Maximizing Retail Value Across Distributed Data
Customer Use Case: Efficiently Maximizing Retail Value Across Distributed Data Warehouse Systems Klaus-Peter Sauer Technical Lead SAP CoE EMEA at Teradata Agenda 1 HEMA Company Background 2 Teradata Overview 3 Why HEMA choose Teradata 4 The Implementation 5 Summary 2 A new store in the Netherlands in 1926 3 Facts Brand awareness in the Netherlands 100% 4.4 million customers per week Daily number of visitors on www.hema.nl: 50.000 HEMA sells a sausage every 3 seconds (10 million a year) One out of three Dutch boys wears HEMA underwear One out of five Dutch women wears HEMA bra 4 Distinguished style This is one of our strongest USP’s Together with low price and high quality 5 Formats High traffic XL AA / D HEMA international 6 6 Hema.nl 7 7 Agenda 1 HEMA Company Background 2 Teradata Overview 3 Why HEMA choose Teradata 4 The Implementation 5 Summary 8 Teradata – Company Overview Teradata Corporation Founded in 1979 2010 Revenue: $1,936M 8,000 Associates in 70 countries Global Leader in Enterprise Data Warehousing 2011 Magic Quadrant Data Warehouse DBMS > Independent since Oct 2007 > S&P 500 Member, listed NYSE (TDC) > First TB+PB DWH on Teradata > Database Technology, Analytic Solutions, Consulting Services Since 1999 #1 Position in “Gartner’s Leader’s Quadrant in Data Warehousing” Teradata Key Offerings Teradata DBMS Teradata MPP Platform The Magic Quadrant is copyrighted January 2011 by Gartner, Inc. and is reused with permission. The Magic Quadrant is a graphical representation of a marketplace at and for a specific time period. It depicts Gartner's analysis of how certain vendors measure against criteria for that marketplace, as defined by Gartner. Gartner does not endorse any vendor, product or service depicted in the Magic Quadrant, and does not advise technology users to select only those vendors placed in the "Leaders" quadrant. The Magic Quadrant is intended solely as a research tool, and is not meant to be a specific guide to action. Gartner disclaims all warranties, express or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose. Teradata SAP Partnership Overview Business Objects Partner since 1995 320+ joint customers globally, across industries Teradata Advisory Group Business Objects is included in both BI and Data Integration portfolios SAP NetWeaver Partner since 2004 Teradata is committed to the SAP NetWeaver platform to provide better, seamless integration between SAP applications and Teradata. Teradata certified SAP NetWeaver Interfaces. Teradata SAP integration development lab in San Diego. Teradata CoE SAP to support the field organizations. Teradata SAP Integration Lab EMEA in Prague. Teradata Office at SAP Partner Port Building in Walldorf. Partner Port Building in Walldorf 10 Teradata Virtual Access for SAP Teradata Extract and Load Solution Teradata Supply Chain Accelerator Jan 2008 Nov 2006 > Using Virtual Info Cubes to access data held in Teradata > Easily combine SAP and non-SAP data in BW queries Jun 2007 Oct 2005 SAP NW Integration Products > Use Open Hub to load data from BW to Teradata > Easy extraction of SAP data into Teradata environment > Use Teradata to power SAP Demand Planning Solution > Faster, more frequent planning cycles using greater detail and history Teradata JMS Universal Connector > Teradata Active Data Warehouse for SAP > Message-Bus Integration with SAP NetWeaver PI 11 Agenda 1 HEMA Company Background 2 Teradata Overview 3 Why HEMA choose Teradata 4 The Implementation 5 Summary 12 HEMA Expansion We became Holland’s favorite and we still are! 1926 2 stores 1940 24 stores 1970 95 stores 1985 193 stores 1995 242 stores 2011 +550 stores in the Netherlands, Belgium, Germany, France, Luxembourg 13 13 Expansion is key to HEMA … …but that puts pressure on HEMA supply chain Consequence New formats do not always fit in the current model Local influences (store level) become more important Conclusion: new Supply Chain model is required: Demand driven Based upon local influences Management by Exception Teradata selected to support HEMA strategy: DCM application SAP BW integration 14 Challenges Demand Chain Management New Demand Chain (DCM) application on Teradata chosen as foundation of HEMA’s new Supply Chain model Analysis did show, that most of the data needed to feed the DCM application already stored in SAP BW Potential Data duplication issue raised SAP Business Warehouse Fast data SAP BW volume growth expected Query performance issue with SAP BW on Oracle perceived 15 Strategy and Project Rules Leverage the Teradata DCM investment also to solve SAP BW (Oracle) performance issue Avoid data redundancy - “Single version of the truth” Data scope: Sales and Stock subject area (~50% of SAP BW data) (Re-)Use current SAP BW ETL / Reporting Keep or improve query performance Performance test halfway the project! 16 Agenda 1 HEMA Company Background 2 Teradata Overview 3 Why HEMA choose Teradata 4 The Implementation 5 Summary 17 SAP BW at HEMA usage data About 500 HQ users + Distribution Center Users All shops in all countries(550+) Monday morning peak sizing Sales (per day-article-plant) No receipts Stock (article-week-shop) Remote cube to R/3 (actual stock) Article movements Financial data (pca, cca, sl) used tools 2,5TB+ data at this moment 150+ InfoCubes 1000+ report queries BEX (Web) Analyzer BEX Report Designer BEX Broadcasting 18 18 Implementation in a nutshell 1. Teradata infrastructure implementation and set up 2. Integration Teradata and SAP BW: – Data flows from SAP BW to Teradata via SAP OpenHub and Teradata TELS – Queries get data out of Teradata via Teradata TVAS 3. Implementation Teradata DCM on top of Teradata DW SHS DCM (SAP HEMA Store) Stock / Sales / SAP Retail (ECC 6.0) SAP BW Master Data TVAS Daily replenishment order proposals 19 Teradata DW Teradata Virtual Access Solution TVAS allows SAP BW End-users to run reports against data which is physically stored in Teradata only. TVAS avoids data duplication and ETL implementation. TVAS gives SAP BW End-users high performance access to detailed data in Teradata. TVAS key functionality is a Teradata specific SQL generator. TVAS runs on SAP NetWeaver Java Application Server and supports multiple BW instances including SAP Java load balancing. TVAS supports multiple Teradata systems and Teradata query banding. 20 Reduced Cost Improved Performance Increased Business Value by more fresh and detailed data TVAS Use Cases Illustrative 21 HEMA Solution Architecture Teradata Complements SAP BW 22 Illustrative Step 1: Simplify the Data Model Basic Design Idea – Store once, use many! 23 Illustrative Step 2: Initial Data Load • Load historical info available from 2006 – Sales Data – Stock Data – Master Data • Method: – Export from SAP BW to a Flat File – Import in Teradata with Loader 24 Step 3: Data Mapping SAP BW Virtual Provider to Teradata (TVAS GUI) 25 Step 4: Daily ETL Embedded in existing HEMA/CapGemini environment, use of: – BMC Control-M scheduling ETL – Export : via SAP BW export via Open Hub – Load: via Teradata Load Solution (TELS) and FTP/Teradata loader: load SAP BW data in Teradata Staging Area – Transform: via Teradata SQL: update Data model 26 BW & Teradata in Production Results & Findings Query performance improved significantly Users do not complain (so much) anymore Very stable environment New queries developed to combine SAP and DCM data Previous response time Current timings (average) A < 10 sec 2x faster B 10 < > 60 sec 2x faster C 60 < > 300 sec 10x faster D > 300 sec 24x faster Group 27 27 Agenda 1 HEMA Company Background 2 Teradata Overview 3 Why HEMA choose Teradata 4 The Implementation 5 Summary 28 Implementation Summary No difference for BW End-User Substantial performance improvement Store once, use many Simplified Data Model and structures Implementation with a small team in 4 months Cost savings on storage & maintenance Compare before and after – More users and more usage – More historical data on the system – More data requested in the reports 29 Looking forward Teradata role for HEMA is changing… SAP BW Teradata Commodity reporting Special reporting Large group of users: Stores and Head office Special Head office users only Aggregated data Detailed data Data hub to Teradata Data supply to BW for non SAP data SAP Merchandise and Assortment Planning POS Data Web Data 30 Contact Klaus-Peter Sauer Technical Lead SAP Program Europe – Middle East – Africa Teradata GmbH Altrottstr. 31 69190 Walldorf / Germany Tel: +49 (0) 6227 / 733 511 Mobile: +49 (0) 172 / 8238 665 Fax: +49 (0) 89 / 3221 1974 Klaus-Peter.Sauer@teradata.com Teradata.com