Architecture of Data Warehousing-Explanation of Architecture of Data Warehousing

Architecture of Data Warehousing

In Architecture of Data Warehousing 4 levels are occurs. The fallowing parts declare the data warehouse architecture in detail.

1.Data Sources

2.Bottom Level (Data warehouse Server)

3.Middle Level(OLAP Server)

4.Top Level(Front End Tool)

Diagram Downloaded From Internet- Credit To panoply.io

1.Data Source-

In data source stores all data about that organization, In that include day to day transnational data and also store in the form of word file, video, audio, 2D, 3D, etc. files. That all data not is well formatted data.

E.g.-Complete about organization like training details, sales detail, dept detail, emp detail etc.

The all data not well documented or well structured that reason from that data not gathering any useful information for user.

2.Bottom Level(Data warehouse Server)-

Data warehouse getting only useful information from data source based on data mining knowledge. Data extracted from data sources with use of back end tools like-Data Extraction, Data Cleaning, Transformation.

Bottom Level perform fallowing task on data , In that include-

A. Data Warehouse

B. Metadata repository

C. Data mart

D. Monitoring and administration

A. Data Warehouse-

In Architecture of Data Warehousing -Data Warehouse stage extracted data from operation database or data sources mainly. That contains only relevant information about user expectations. In DW all data are subjected and well formed.

B. Metadata Repository-

Metadata means data about data. Metadata represent root map about data warehouse. In metadata stores all information about data present in data warehouse. In that include-

1.Structure of data warehouse

2.Data names and definitions

3.Methods used for data cleaning information

4.Incoming and outgoing data information

C. Data mart-

It is a subset of data warehouse. In that contains only small slices of data warehouse. In that data mart include specific information about any department or category. Data mart is a small slices of data warehouse.

D. Monitoring and Administration-

In that manages task about security about data, also manages and control data task, query executions. The all observation and rules and regulation applying from here to all data.

3. Middle Level-(OLAP Server)-

OLAP represent multidimensional data to user from data warehouse or data mart. It represent data with different dimension fro processing fast. They provide speed and less time for getting data from large data warehouse.

It allows accessing information fast, reliable and consistent manner. In that use OLAP types for accessing data like-

1.Relational OLAP(ROLAP)

2.Multi-dimensional OLAP(MOLAP)

3.Hybrid OLAP(HOLAP)

with specialized SQL queries. The OLAP method uses different technique for data dealing and display like-

1.Roll up -Roll Down

2.Slice and Dice

3.Drill up and Drill Down

4.Pivoting

That all factor are easy to use and useful for getting data fast to next stage.

4.Top Level-(Front End Tools)-

It is a front end tool to use showing data to user.

1.In that stage use different reporting/Query tools for display data.

2.The analyzing tools are prepare for chart based analysis and show result to user.

3.Data Mining tools are used for getting knowledgeable information from big data and convert that data into one pattern and show result to user with use of GUI tools.

Explanation :

The architecture of a data warehouse defines how data is collected, stored, managed, and accessed for analysis and decision-making. It serves as the foundation for organizing large volumes of data from various sources into a unified system that supports business intelligence and analytical processes. Generally, data warehouse architecture is structured in three layers: Data Source Layer, Data Storage Layer, and Presentation Layer, supported by additional components such as metadata and data staging areas.

1. Data Source Layer:

This layer includes all operational systems and external data sources that provide raw data for the warehouse. These sources may include transactional databases, CRM systems, ERP applications, flat files, or external APIs. Data from these systems is extracted using ETL (Extract, Transform, Load) tools. The extraction process gathers data, transformation ensures data consistency and accuracy, and loading transfers it into the data warehouse for storage.

2. Data Storage Layer:

Also known as the Data Warehouse Database, this layer is the central repository where cleaned and transformed data is stored. The storage is typically designed using a relational or multidimensional model. The data is organized into fact and dimension tables that support efficient querying and analysis. In many architectures, a staging area is used temporarily to process data before loading it into the warehouse. Additionally, data marts—smaller, subject-oriented sections of the data warehouse—may be created to cater to specific departments like sales or finance.

3. Presentation Layer:

This layer provides access to end users for reporting, querying, and analysis. It includes tools such as dashboards, OLAP cubes, and data visualization applications that allow users to explore data and gain insights. This layer transforms raw data into meaningful information through summaries, charts, and performance indicators.

Supporting Components:

Metadata acts as the “data about data,” defining source details, transformations, and structures. Data mining and analytical tools further enhance insight extraction.

Read More-

1. What Is Data Warehouse
2. Applications of Data Warehouse, Types Of Data Warehouse
3. Architecture of Data Warehousing
4. Difference Between OLTP And OLAP
5. Python Notes