Data Warehouse, Data Mart, ETL, OLAP, OLTP, Dimensions, Schema, Star & snowflake schema

1.             Definition of data warehousing?

ü   Data warehouse is a Subject oriented, Integrated, Time variant, Non volatile collection of data in support of management's decision making process.

                                                                                     
                                   

Subject Oriented

Data warehouses are designed to help you analyze data. For example, to learn more about your company's sales data, you can build a warehouse that concentrates on sales. Using this warehouse, you can answer questions like "Who was our best customer for this item last year?" This ability to define a data warehouse by subject matter, sales in this case makes the data warehouse subject oriented.

Integrated

Integration is closely related to subject orientation. Data warehouses must put data from disparate sources into a consistent format. They must resolve such problems as naming conflicts and inconsistencies among units of measure. When they achieve this, they are said to be integrated.

Nonvolatile

Nonvolatile means that, once entered into the warehouse, data should not change. This is logical because the purpose of a warehouse is to enable you to analyze what has occurred.

Time Variant

In order to discover trends in business, analysts need large amounts of data. This is very much in contrast to online transaction processing (OLTP) systems, where performance requirements demand that historical data be moved to an archive. A data warehouse's focus on change over time is what is meant by the term time variant.

2.             How many stages in Data warehousing?

Data warehouse generally includes two stages

ü   ETL

ü   Report Generation

ETL

Short for extract, transform, load, three database functions that are combined into one tool

·                     Extract -- the process of reading data from a source database.

·                     Transform -- the process of converting the extracted data from its previous form into required form

·                     Load -- the process of writing the data into the target database.

ETL is used to migrate data from one database to another, to form data marts and data warehouses and also to convert databases from one format to another format.

It is used to retrieve the data from various operational databases and is transformed into useful information and finally loaded into Data warehousing system.

1 INFORMATICA

2 ABINITO

3 DATASTAGE

4. BODI

5 ORACLE WAREHOUSE BUILDERS

Report generation

          In report generation, OLAP is used (i.e.) online analytical processing. It is a set of specification which allows the client applications in retrieving the data for analytical processing.

It is a specialized tool that sits between a database and user in order to provide various analyses of the data stored in the database.

OLAP Tool is a reporting tool which generates the reports that are useful for Decision support for top level management.

1.             Business Objects

2.             Cognos

3.             Micro strategy

4.             Hyperion

5.             Oracle Express

6.    Microsoft Analysis Services

·                     Different Between OLTP and OLAP

	OLTP	OLAP
1	Application Oriented (e.g., purchase order it is functionality of an application)	Subject Oriented (subject in the sense customer, product, item, time)
2	Used to run business	Used to analyze business
3	Detailed data	Summarized data
4	Repetitive access	Ad-hoc access
5	Few Records accessed at a time (tens), simple query	Large volumes accessed at a time(millions), complex query
6	Small database	Large Database
7	Current data	Historical data
8	Clerical User	Knowledge User
9	Row by Row Loading	Bulk Loading
10	Time invariant	Time variant
11	Normalized data	De-normalized data
12	E – R schema	Star schema

3.             What are the types of data warehousing?

EDW (Enterprise data warehousing)

ü   It provides a central database for decision support throughout the enterprise

ü   It is a collection of DATAMARTS

DATAMART

ü   It is a subset of Data warehousing

ü   It is a subject oriented database which supports the needs of individuals depts. in an organizations

ü   It is called high performance query structure

ü   It supports particular line of business like sales, marketing etc.

ODS (Operational data store)

ü   It is defined as an integrated view of operational database designed to support operational monitoring

ü   It is a collection of operational data sources designed to support Transaction processing

ü   Data is refreshed near real-time and used for business activity

ü   It is an intermediate between the OLTP and OLAP which helps to create an instance reports

4.    What are the modeling involved in Data Warehouse Architecture?

5.    What are the types of Approach in DWH?

Bottom up approach: first we need to develop data mart then we integrate these data mart into EDW

Top down approach: first we need to develop EDW then form that EDW we develop data mart

Bottom up

OLTP             ETL                Data mart                   DWH        OLAP

Top down

OLTP             ETL                DWH                Data mart             OLAP

Top down

ü   Cost of initial planning & design is high

ü   Takes longer duration of more than an year

Bottom up

ü   Planning & Designing the Data Marts without waiting for the Global warehouse design

ü   Immediate results from the data marts

ü   Tends to take less time to implement

ü   Errors in critical modules are detected earlier.

ü   Benefits are realized in the early phases.

ü   It is a Best Approach

Data Modeling Types:

ü   Conceptual Data Modeling

ü   Logical Data Modeling

ü   Physical Data Modeling

ü   Dimensional Data Modeling

1. Conceptual Data Modeling

ü   Conceptual data model includes all major entities and relationships and does not contain much detailed level of information about attributes and is often used in the INITIAL PLANNING PHASE

ü   Conceptual data model is created by gathering business requirements from various sources like business documents, discussion with functional teams, business analysts, smart management experts and end users who do the reporting on the database. Data modelers create conceptual data model and forward that model to functional team for their review.

ü   Conceptual data modeling gives an idea to the functional and technical team about how business requirements would be projected in the logical data model.

2. Logical Data Modeling

ü   This is the actual implementation and extension of a conceptual data model. Logical data model includes all required entities, attributes, key groups, and relationships that represent business information and define business rules.

3. Physical Data Modeling

ü   Physical data model includes all required tables, columns, relationships, database properties for the physical implementation of databases. Database performance, indexing strategy, physical storage and demoralization are important parameters of a physical model.

Logical vs. Physical Data Modeling

Logical Data Model	Physical Data Model
Represents business information and defines business rules	Represents the physical implementation of the model in a database.
Entity	Table
Attribute	Column
Primary Key	Primary Key Constraint
Alternate Key	Unique Constraint or Unique Index
Inversion Key Entry	Non Unique Index
Rule	Check Constraint, Default Value
Relationship	Foreign Key
Definition	Comment

Dimensional Data Modeling

ü   Dimension model consists of fact and dimension tables

ü   It is an approach to develop the schema DB designs

Types of Dimensional modeling

ü   Star schema

ü   Snow flake schema

ü   Star flake schema (or) Hybrid schema

ü   Multi star schema

What is Star Schema?

ü   The Star Schema Logical database design which contains a centrally located fact table surrounded by at least one or more dimension tables

ü   Since the database design looks like a star, hence it is called star schema db

ü   The Dimension table contains Primary keys and the textual descriptions

ü   It contain de-normalized business information

ü   A Fact table contains a composite key and measures

ü   The measure are of types of key performance indicators which are used to evaluate the enterprise performance in the form of success and failure

ü   Eg: Total revenue , Product sale , Discount given, no of customers

ü   To generate meaningful report the report should contain at least one dimension and one fact table

The advantage of star schema

ü   Less number of joins

ü   Improve query performance

ü   Slicing down

ü   Easy understanding of data.

Disadvantage:

ü    Require more storage space

Example of Star Schema:

Snowflake Schema

ü   In star schema, If the dimension tables are spitted into one or more dimension tables

ü   The de-normalized dimension tables are spitted into a normalized dimension table

Example of Snowflake Schema:

ü   In Snowflake schema, the example diagram shown below has 4 dimension tables, 4 lookup tables and 1 fact table. The reason is that hierarchies (category, branch, state, and month) are being broken out of the dimension tables (PRODUCT, ORGANIZATION, LOCATION, and TIME) respectively and separately.

ü   It increases the number of joins and poor performance in retrieval of data.

ü   In few organizations, they try to normalize the dimension tables to save space.

ü   Since dimension tables hold less space snow flake schema approach may be avoided.

ü   Bit map indexes cannot be effectively utilized

Important aspects of Star Schema & Snow Flake Schema 

ü   In a star schema every dimension will have a primary key.

ü   In a star schema, a dimension table will not have any parent table.

ü   Whereas in a snow flake schema, a dimension table will have one or more parent tables.

ü   Hierarchies for the dimensions are stored in the dimensional table itself in star schema.

ü   Whereas hierarchies are broken into separate tables in snow flake schema. These hierarchies help to drill down the data from topmost hierarchies to the lowermost hierarchies.

Star flake schema (or) Hybrid Schema

ü   Hybrid schema is a combination of Star and Snowflake schema

Multi Star schema

ü   Multiple fact tables sharing a set of dimension tables

ü   Confirmed Dimensions are nothing but Reusable Dimensions.

ü   The dimensions which u r using multiple times or in multiple data marts.

ü   Those are common in different data marts

Measure Types (or) Types of Facts

·                     Additive - Measures that can be summed up across all dimensions.

o                  Ex: Sales Revenue

·                     Semi Additive - Measures that can be summed up across few dimensions and not with others

o                  Ex: Current Balance

·                     Non Additive - Measures that cannot be summed up across any of the dimensions.

o                  Ex: Student attendance

Surrogate Key

ü   Joins between fact and dimension tables should be based on surrogate keys

ü   Users should not obtain any information by looking at these keys

ü   These keys should be simple integers

A sample data warehouse schema

WHY NEED STAGING AREA FOR DWH?

ü   Staging area needs to clean operational data before loading into data warehouse.

ü   Cleaning in the sense your merging data which comes from different source.

ü   It’s the area where most of the ETL is done

Data Cleansing

ü   It is used  to remove duplications

ü   It is used to correct wrong email addresses

ü   It is used to identify missing data

ü   It used to convert the data types

ü   It is used to capitalize name & addresses.

Types of Dimensions:

There are three types of Dimensions

ü   Confirmed Dimensions

ü   Junk Dimensions Garbage Dimension

ü   Degenerative Dimensions

ü   Slowly changing Dimensions

Garbage Dimension or Junk Dimension

ü   Confirmed is something which can be shared by multiple Fact Tables or multiple Data Marts.

ü   Junk Dimensions is grouping flagged values

ü   Degenerative Dimension is something dimensional in nature but exist fact table.(Invoice No)

            Which is neither fact nor strictly dimension attributes. These are useful for some kind of analysis. These are kept as attributes in fact table called degenerated dimension

Degenerate dimension: A column of the key section of the fact table that does not have the associated dimension table but used for reporting and analysis, such column is called degenerate dimension or line item dimension.

For ex, we have a fact table with customer_id, product_id, branch_id, employee_id, bill_no, and date in key section and price, quantity, amount in measure section. In this fact table, bill_no from key section is a single value; it has no associated dimension table. Instead of creating a

Separate dimension table for that single value, we can Include it in fact table to improve performance. SO here the column, bill_no is a degenerate dimension or line item dimension.