DATA WAREHOUSING

Basics Concepts

People Making Technology Work

Agenda
Evolution of DWH Why should we consider Data Warehousing solutions ? Definition of Data Warehouse Characteristics of DWH Difference between DWs and OLTP DWH Life Cycle DWH Architecture Dimensional Data Modeling Star Schema Design Fact Table Fact Granularity Dimension Tables Snowflake Schema Design Important aspects of Star Schema & Snow Flake Schema Data Acquisition (ETL) ETL Concepts

Evolution of DWH

Traditional approaches to computer system design during 1980s

Not optimized for analysis and reporting Company wide reporting couldnt be supported from a single system For developing reports often required writing specific computer programs which was slow and expensive

Why should we consider Data Warehousing solutions ?

When users are requesting access to a large amount of

historical information for reporting purposes, you should

strongly consider a warehouse or mart. The user will benefit when the information is organized in an efficient manner for this type of access.

Def . Data Warehousing

DWH is type of relational data base system specially

designed for query analysis processing rather than

transactional processing.

The DWH systems are also called as Historical Dbs,

Read only Dbs, Integrated Dbs, Decision Supporting System, Executive info System, Business Info System.

Characteristics of DWH

Subject Oriented Non Volatile Integrated

Time Variant

Differences..
DWH database (OLAP)
Designed for analysis of business measures by category and attributes. Optimized for bulk loads and large, complex, unpredictable queries that access many rows per table. Loaded with consistent, valid data; requires no real time validation. Supports few concurrent users relative to OLTP.

OLTP database
Designed for real time business operations.
Optimized for a common set of transactions, usually adding or retrieving a single row at a time per table. Optimized for validation of incoming data during transactions; uses validation data tables. Supports thousands of concurrent users.

OLAP Database (OLAP)

Multidimensional Database Structures Index - Many

OLTP Database
Normalized Data Structures Index - Few

Joins - Few
Aggregated Data - More No. of users - Few Periodic update of data Huge volumes of data

Joins - Many
Aggregate Data - Few No. of users - More Data Modification More Small volumes of data

DWH Life Cycle

Business Analyst Data Modular

ETL Developer

Report Developer

Testing

DWH Architecture
Three common architectures are:

DWH Architecture (Basic)

DWH Architecture (With a staging area) DWH Architecture (With a staging area and data marts)

DWH Architecture (Basic)

DWH Architecture (with a staging area)

DWH Architecture (with a staging area and data marts)

Dimensional Data Modeling

To develop a Star Schema design a Data Modeler follows dimensional modeling design aspect. Dimensional modeling is a 3 stage process

Conceptual modeling Logical Modeling Physical Modeling

Before start implementing the schema design a Data modeler should understand the following process Understand the clients Business requirements Understand the grain of fact Designing of the Dimension tables Designing of the Fact tables

Example of Dimensional Data Model (Star Schema Design)

Fact Table
Contain numeric measures of the business Contains facts and connected to dimensions two types of columns facts or measures foreign keys to dimension tables May contain date-stamped data A fact table might contain either detail level facts or facts that have been aggregated

Steps in designing Fact Table

Identify a business process for analysis(like sales). Identify measures or facts (sales dollar). Identify dimensions for facts(product dimension, location dimension, time dimension, organization dimension). List the columns that describe each dimension.(region name, branch name, region name). Determine the lowest level of summary in a fact table(sales dollar).

Types of Facts (Measures)

Additive - Measures that can be added across all dimensions. Semi Additive - Measures that can be added across few dimensions and not with others. Non Additive - Measures that cannot be added across all dimensions.

In the example, sales fact table is connected to dimensions location, product, time and organization. Measure "Sales Dollar" in sales fact table can be added across all dimensions independently or in a combined manner which is explained below. Sales Dollar value for a particular product Sales Dollar value for a product in a location Sales Dollar value for a product in a year within a location Sales Dollar value for a product in a year within a location sold or serviced by an employee

Fact Granularity
A fact table maintains a numerical info It is defined as the level at which fact info/- is stored. The level is determined by dimensional table. Year? Quarter? Month? Week? Day?

Dimension Tables
Contain textual information that represents attributes of the business Contain relatively static data Are joined to fact table through a foreign key reference Are usually smaller than fact tables

Example of Location Dimension

Location Dimension
Location Dimension

Location Dimension Id

Country Name

State Name

County Name

City Name

Date Time Stamp

USA

New York

Shelby

Manhattan

1/1/2005 11:23:31 AM 1/1/2005 11:23:31 AM 1/1/2005 11:23:31 AM 1/1/2005 11:23:31 AM

USA

Florida

Jefferson

Panama City

USA

California

Montgomery

San Hose

USA

New Jersey

Hudson

Jersey City

Star Schema Design benefits

Easy for users to understand

Fast response to queries

Support multi dimensional analysis Supported by many front end tools

Snowflake Schema Design

Dimension table hierarchies are broken into simpler tables In few organizations, they try to normalize the dimension tables to save space Both Fact and Dimensional tables are Normalized Increases the number of joins and poor performance in retrieval of data May become large and unmanageable Degrades query performance

Example of Snowflake Schema

Important aspects of Star Schema & Snow Flake Schema

In a star schema every dimension will have a primary key. In a star schema, a dimension table will not have any parent table. Whereas in a snow flake schema, a dimension table will have one or more parent tables. Hierarchies for the dimensions are stored in the dimensional table itself in star schema. Whereas hierarchies are broken into separate tables in snow flake schema. These hierarchies helps to drill down the data from topmost hierarchies to the lowermost hierarchies.

Data Acquisition
It is the process of extracting the relevant business info/- from the different source systems transforming the data from one format into an another format, integrating the data in to homogeneous format and loading the data in to a warehouse database. Data Extraction (E) Data Transformation (T) Data Loading (L)

Sample ETL Process Flow

ETL Process
The ETL Process having the following basic steps Is mapping the data between source systems and target database Is cleansing of source data in staging area Is transforming cleansed source data and then loading into the target system

 Source System A database, application, file, or other storage facility from which the data in a data warehouse is derived. Mapping The definition of the relationship and data flow between source and target objects. Staging Area A place where data is processed before entering the warehouse. Cleansing The process of resolving inconsistencies and fixing the anomalies in source data, typically as part of the ETL process.

 Transformation The process of manipulating data. Any manipulation beyond copying is a transformation. Examples include cleansing, aggregating, and integrating data from multiple sources. Transportation The process of moving copied or transformed data from a source to a data warehouse. Target System A database, application, file, or other storage facility to which the "transformed source data" is loaded in a data warehouse.

Thank You !!!