Asuri Saranathan BEST PRACTICES IN DATA WAREHOUSING AND

Asuri Saranathan BEST PRACTICES IN DATA WAREHOUSING AND BUSINESS INTELLIGENCE

Agenda Introduction Best Practices – Over View Deep Dive Conclusion Q&A

Introduction

Speaker Holds Bachelor degree in Physics and Electrical and Electronics Engineering. Over 26 years of Experience in Power System and Information Technology field. Has built several large scale applications including online CRM for multinationals. Has managed several Data Warehousing and BI projects for Direct Marketing, Manufacturing and Auto finance verticals. Functioned as Solution Architect for Data warehousing and reporting projects. ISO auditor Certified Bullet Proof Manager from Crest. Comm USA.

Best Practices- Overview

What are Best Practices? Is it a technology? Is it application of a set of best tools available in the market? Is it a Framework?

Best Practices Definition A framework of a set of processes or method that exhibits achievement of specific results in a specific manner over a sustained period of time. The framework should have certain characteristics in that they should be repeatable.

Do Best Practices Evolve? Yes they do. Because of Innovation Changes in Technology Changes in Law or Governance Structure. Expectations, Values , Knowledge or other that makes the practice outdated or inappropriate.

Where can it be Applied? Practically in all fields.

How do we apply Best Practices to Data Warehousing and Business Intelligence?

Data Warehouse - Definition In an elementary form , it is the collection of key information that can be used by the business users to become more profitable. But Is this definition sufficient ? We need much more precise definition of what a data ware house is.

What is a Data Warehouse? A Data warehouse is the Data ( Meta / Fact / Dimension/ Aggregation) and The Process Managers ( Load / Warehouse / Query) That make information available , enabling the user to make informed decisions.

Deep Dive

What is the Challenge? Business is never Static. And so is Data Warehouse. In order to respond to today’s requirement for instant access to corporate information , the data warehouse should be designed to respond to this need in a optimal way. Business itself probably not aware of what information is required in the future. This requires a fundamentally different approach than the traditional waterfall method of software development for the Data warehouse.

Experience so far… Most Enterprise Data Warehousing projects tend to have development cycle of between 18 – 24 months from start to finish. Justification of this investment is substantial. Businesses would prefer a better approach to justify the investment.

What should be done? Focus on Business Requirements A clear understanding of what is short term and long term requirement of the data warehouse. An Architecture design that would evolve. Identification of quick win that delivers business benefit in the first build.

Scalability for Growth Scalability means ability of the underlying Hardware and Software to support increasing demands over a period of time.

Horizontal Scalability High Speed Network CPU CPU RAM RAM DB DB Multiple servers are connected thru a network and use the data partitioning feature of the Database to tie the CPUs together.

Data Warehouse Environment Staging Area Source Systems Data warehouse (System of Record) Full History in 3 rd Normal Form No User Access Summary Area Full History User Access Analytical Area User Access Data Mart

Strategy Metadata Mgmt • Architecture • Integration • Control • Delivery Document / Content Mgmt • Acquisition & Storage • Backup & Recovery • Content • Retrieval • Retention DWH / BI • Architecture • Implementation • Training and Support • Tuning Data Architecture • Entp. DM • Value Chain Data Quality • Spec • Analysis • Measurement • Improvement Data Governance Reference and MDM • External & Internal Code • Customer Data • Product data Development • Analysis • Data Modeling • DB Design • Implementation Data Operations • Acquisition • Recovery • Tuning • Purging Data Security • Standards • Classification • Administration • Authentication • Auditing

Environment Organization & Culture Technology Activities Goals & Objectives Practices & Techniques Deliverables Roles & Responsibilities

Architecture Requirements Must recognize change as a constant Take incremental development approach Existing applications must continue to work Need to allow more data and new types of data to be added

High Level Remember the different “worlds” On-line transaction processing (OLTP) Business intelligence systems (BIS) Users are different Data content is different Data structures are different Architecture & methodology must be different

Best Practice #1 Ø Use a Data model that is optimized for Information retrieval dimensional model denormalized hybrid approach DW Architecture Best Practices 19 March 2018 24

Best Practice #2 Ø Carefully design the data acquisition and cleansing processes for your DW Ensure the data is processed efficiently and accurately Consider acquiring ETL and Data Cleansing tools Use them well! DW Architecture Best Practices 19 March 2018 25

Best Practice #3 Ø Design a metadata architecture that allows sharing of metadata between components of your DW consider metadata standards such as OMG’s Common Warehouse Metamodel (CWM) DW Architecture Best Practices 19 March 2018 26

Best Practice #4 Ø Take an approach that consolidates data into ‘a single version of the truth’ Data Warehouse Bus conformed dimensions & facts OR? DW Architecture Best Practices 19 March 2018 27

Best Practice #5 Ø Consider implementing an ODS only when information retrieval requirements are near the bottom of the data abstraction pyramid and/or when there are multiple operational sources that need to be accessed Must ensure that the data model is integrated, not just consolidated May consider 3 NF data model Avoid at all costs a ‘data dumping ground’ DW Architecture Best Practices 19 March 2018 28

Pitfalls to be Avoided Engagement of Non-BI Manger in a BI delivery Project. Trying to please the client and the user community. Expecting the Service Provider to own the Project completely. Bringing the Solution Architect half way into the project. Allowing the Business Users to drive the Data Model. Not having the right people with right skills in tool selection process. Expecting the contractor to deliver all that they presented. Over dependency on the Service provider or contractor in execution. Assuming that the Data quality will be handled somehow. Assuming that the Data warehouse project is over once it is deployed.

Data Warehouse Architecture Best Practices Thank You