Data Warehouses

In this step, we’ll explore the difference in design and purpose between a data warehouse and a database.

Data Lake vs Data Warehouse

Last week, you learned about Azure Data Lake Store, which is traditionally used to store large volumes of raw data before it’s transformed or acted upon. This is often where data might start its journey.

Data warehouses are also large stores, but their purpose is predefined and tends to be a store of historic data for reporting purposes. They are special forms of relational databases.

Fact & Dimension Tables

Data warehouses commonly take a star schema or snowflake schema. The idea is that there’s a central fact table that holds metrics or facts about a business process which can act as foreign keys, and dimension tables surround it relating via their primary or candidate keys to give further context to the facts.

Reporting from historical data

One of the main values of data warehousing is establishing trends in the data.

New SQL commands

• GROUP BY – an organisational command that sorts your results through the use of a column.

Data Warehousing

Data warehousing is a specific database workload for large volumes of historical data that can be used for reporting and analysis. If you’ve never worked with a star schema data warehouse, you may find it useful to explore more. It will be expanded on in later steps.

Note: Why not explore more about the new SQL command, GROUP BY introduced in this step? Find out more about it at W3School’s SQL Group By page.