Data Warehouse approaches with Dynamics AX
0 likes6,025 views
Dynamics AX의 BI 구축을 위해 필요한 Data Warehouse 내용입니다. • What is a Data Warehouse • Data Warehouse Approaches • Why Invest in a Data Warehouse • Getting Started • BI Models • BI Solutions
Ad
More Related Content
What's hot (20)
Viewers also liked (20)
Ad
Similar to Data Warehouse approaches with Dynamics AX (20)
Ad
More from Alvin You (8)
Recently uploaded (20)
Data Warehouse approaches with Dynamics AX
- 1. Data Warehouse Approaches with Dynamics AX UBAX12 Joel S. Pietrantozzi Executive Vice President Client Strategy Group CLIENT STRATEGY GROUP
- 2. Agenda • What is a Data Warehouse • Data Warehouse Approaches • Why Invest in a Data Warehouse • Getting Started • BI Models • BI Solutions
- 3. Introduction • Joel S. Pietrantozzi – Executive Vice President, Client Strategy Group – O: 216.524.2574 – Email: [email protected] CLIENT STRATEGY GROUP
- 4. Introduction • Client Strategy Group – Revive • Implementation Turnaround • AX Performance Tuning – Enhance • Business Intelligence • Increased Value – Upgrade • Strategy & Planning • Implementation CLIENT STRATEGY GROUP
- 5. AXUG Premier Partner AXUG Training Academy Classes 1. AX 2012 – Upgrade your code 2. AX 2012 – Upgrade your data 3. AX 2012 – Understanding the Data Model 4. AX2012 – Understanding the Security Model 5. AX 2012 – Performance Optimization 6. AX 2012 – Managing your Environment 7. AX 2009 – Performance Optimization
- 6. WHAT IS A DATA WAREHOUSE?
- 7. What is a Data Warehouse? • Means different things to different people • Complexity factor – Does not have to include ETL • Consider Replication for reporting • Usually fed from many different data sources • Contains a large amount of current and historic data • Allows for flexible reporting, trending and analysis…
- 8. What is a Data Warehouse? • Can simplify the complexity of ad hoc reporting/analysis • Bottom line: – Does it meet reporting/analysis needs – Is the data consistent – Is it flexible in its design? – Can it grow with the organization
- 10. Data Warehouse Approaches (Storage) • Two major approaches – Dimensional – Ralph Kimball • Facts and dimensions • Typically easier to use and understand • Can be complex to maintain/change – Relational – Bill Inmon • Database normalization • Straightforward to add data • Schema paralysis
- 11. Data Warehouse Approaches (Design) • Bottom-up – Result of initial business-oriented top-down analysis – Data marts are created to provide reporting and analysis for specific business processes – Separation of data into segmented data marts – Allows for creation of smaller, less-complex models
- 12. Data Warehouse Approaches (Design) • Top-Down – Data is stored at the lowest level of detail • Atomic – Generates consistent view of data – Creation of new data marts is relatively simple – Up-front cost can be higher than the bottom-up approach
- 13. Data Warehouse Approaches (Design) • Hybrid – Often resemble a hub and spoke architecture – Legacy, ERP and other production systems can feed • PLC line data – Operational data store + cube set
- 14. WHY INVEST IN A DATA WAREHOUSE
- 15. Why invest in a Data Warehouse? • ERP systems are designed for transactions, not reporting. – Building reports can lead to system performance degradation and can be quite complex. – Report development is usually an IT Department task. • Business Intelligence systems are designed and optimized for reporting and analysis. – Data is cleansed. – Data can be pulled from several different sources for true enterprise analysis. • A business intelligence system is company specific. – It is designed based on requirements.
- 16. Why invest in a Data Warehouse? • Provides a “common truth” for a company’s information. • Provides flexibility for dynamic, proactive analysis as opposed to a static view of information. • Allows users to create analysis/reports pertinent to their needs. • The need for similar reports is eliminated.
- 17. Why invest in a Data Warehouse? • Should remove reporting performance hits from Production AX • Multi-dimensional structure in cubes • Eliminates the need for “Rogue” applications • The need for similar reports is eliminated.
- 18. GETTING STARTED
- 19. Getting Started….. • DW topics to consider: – Data Latency Requirements • Operational Reports (Live…picking tickets, labels, etc.) • Business Reporting (Near Live... open orders, etc.) • Analytical Reporting (Day-1… sales analysis, etc.) – Identify Measures & Dimensions by Functional Area(s) – Cross Functional Data Analysis – Change Management Flexibility (external data, new requirements)
- 20. Getting Started….. – How many production data sources? • What is the authoritative data from overlapping production systems? – Don’t let Reports become the ‘authoritative data source’ • Ex. Allocations – should be setup in AX instead of external cubes or reports • Maintenance & Security become on-going issues – Determine Enterprise Definitions for Reporting • How are discounts and returns reported? • How is margin calculated? Yield?
- 21. Front End Options • DW Design should be FE agnostic – Don’t determine DW solution based on ‘pretty’ FE • Transactional Reports – Reporting Services Reports – Excel Worksheet – Management Reporter – Third Party • Analytical Reports – Reporting Services Reports – KPIs – Excel Worksheet – Third Party
- 22. (Some) Excel BIFE Issues • Excel is (almost) everywhere • Usage in even large enterprises is common • Let’s face it: – Powerful – Easy to learn – Embedded – Quick • However, it can be: – Manual – User Error prone – Historical data refresh issues – Size limitations
- 23. Cube Overview • Cubes – Multidimensional data structure • Non-transactional – Cubes contain pre-aggregated data pivoted at the intersection of the dimension keys • Aggregation provide significant speed – Can contain data from one or more fact tables • Different levels of aggregation can be confusing • Consider separating measure groups into different cubes
- 24. Cube Overview • Fact Tables – Lowest level of grain of source data, rolled up into aggregations in SSAS stored in cubes – The quantitative part (measures) of the OLAP analysis – 1 or more required per cube – Tend to be fairly narrow but long tables
- 25. Cube Overview • Dimensions – This is the qualitative piece of the OLAP analysis – Dimensions can (and should) be shared • Time & Territory are examples – Hierarchies and levels are created to provide higher level groupings • Time – Day, Month, Quarter, Year – The relationships that are defined between dimensions and measure groups in a cube determine how the data in the cube is “sliced”
- 26. Business Intelligence Options • Native Dynamics AX Tools • SQL Server stack • Third Party
- 27. Third Party BI Solutions • Perform a through Evaluation & Selection process based on your reporting and analysis requirements. – How do they load historical and external data? • Authoritative data conflicts? – What is the toolset for change management? – What FE Tools are available? – What is the licensing structure? Maintenance? – Implementation estimate & schedule?
- 28. AX 2012 BI Considerations • MorphX reports deprecated • All Dynamics AX 2012 reports have been rewritten to (AX)RS • Utilize Visual Studio 2010 for report development • External/Historical Data Requirements – Conversion – Storage – Non-SQL Data Sources – IDMF (Intelligent Data Mgmt Framework)
- 29. BI MODELS
- 30. BI Models • All-In-One Role Centers Database Engine (AX)RS Reporting Cubes
- 31. BI Models: All-In-One Dynamics AX 2012 KPIs Cubes Reporting Database Engine
- 32. BI Models: Replication Dynamics AX 2012 KPIs Cubes Reporting Database Engine Dynamics AX 2012 KPIs Cubes Reporting Database Engine Replication
- 33. BI Models: External DW Dynamics AX 2012 KPIs Cubes Reporting Database Engine Data Warehouse KPIs Cubes Reporting Database Engine SSIS Non-AX DS
- 34. BI SOLUTIONS
- 35. Cubes Available (AX 2012) • Accounts payable cube • Accounts receivable cube • Customer relationship management cube • Environmental sustainability cube • Expense management cube • General ledger cube • Production cube • Project accounting cube • Purchase cube • Sales cube • Workflow cube
- 36. Planning and Architecture Considerations • Host the OLAP database on a different server from the OLTP server • Security for cubes is set up separately from security for Dynamics AX via roles in Analysis Services • Security for cubes is not synchronized with security for Dynamics AX • How often should the cubes be processed? • Do you plan to create custom cubes?
- 37. Which one? • Transactional volume • Hardware/Infrastructure • Legacy/Other systems • Staff/Partner skillset
- 38. Best Practices • Acquire a business sponsor • Start “small” • Acquire expertise (hire, grow, contract) • Create a solid design – Flexible • Ensure data quality – ETL • “Don’t put the cart before the horse” • “Don’t put the FE before your data”
- 39. External Data Warehouse Model
- 40. Continue the Conversation Online user community for knowledge sharing: • http://community.AXUG.com AXUG events: • http://www.AXUG.com - Webinars and Special Interest Groups (SIGs) • Social Media #AXUG #CONV13 #MSDYNAX And don’t forget to complete your session surveys on the Convergence website, your feedback is appreciated