Sie sind auf Seite 1von 9

Data Warehousing Concepts

Introduction
 Data from multiple sources which would be otherwise hard to query are integrated and loaded
into centralized location
 Before loading data into Warehouse, the data is cleansed and checked for inconsistencies
 Standardizes data across organization
 Since they are separate from OLTP, they reduce the load on OLTP systems. Imagine if there was
no warehouse then the users would be using OLTP for querying, hence slowing down the
system.

What is Data Warehouse?


 A data warehouse is a central location where consolidated data from multiple locations are
stored
 The end user accesses it whenever needed
 It is not refreshed as and when new data is generated But specific batch run timelines are
decided by business.
 Definition:

Data Warehouse Architecture


Why should one Learn Data Warehousing?
 There was a time when business decisions were taken based on Experience
 But experience alone is not sufficient to succeed in business today and can’t be stated as sole
criteria to take decisions
 Hence, analysis of data comes into picture to take effective decisions. Companies are moving
towards taking decisions based on data as they are far more accurate and add a lot more value
 Data Warehouse comes into picture to achieve above aim which helps in storing data to be used
for analytical purpose.

OLTP vs OLAP

OLTP: On-Line Transactional Processing

OLAP: On-Line Analytical Processing


What is Granularity of Data stores in DW?

In Data Warehouse, Data is stored at high Level or in summarized manner, hence, with Low Granularity
in DW as compared to OLTP.

Whereas, OLTP stores low level data i.e., with minute details, hence, OLTP contains high granularity
data.
Data Mart

Data Mart vs Data Warehouse (Difference)

Types of Data Mart


1. Independent Data Mart

2. Dependent Data Mart


3. Hybrid Data Mart

Type of DM Approach
1. Top Down Approach
Data Warehouse are built first and then Data Marts

2. Bottom Up Approach
Data Marts are built prior to Data Warehouse
Operational Data Source (ODS)

Example of ODS
Dimensions and Facts
Dimensions
Facts

Example
Walgreens – Data Warehouse
Data Marts: Item, Merchandising, Sales, Marketing, Inventory, Logistics, Employee Repository

Centralized Store Ordering – Data Warehouse

Dimensions:
o Item, ItemParam, Vendor, Location, SKU, SKUParam, Other SKU Tables, Network,
Sourcing
Facts:
o Sales data pulled from Data Warehouse
o Forecast Data
o On-Hand and Open Orders

Das könnte Ihnen auch gefallen