What is a Data Warehouse

What is data warehousing or what is a business intelligence engineer doing, and why are they using a data warehouse?

To use the analogy to a physical retail-type warehouse, you want to sell very structured products in the most efficient way to your customers. In a data warehouse (DWH) you have typically structured data and optimised them for business users to query. If you dig a little deeper, you offload data from the trucks in the back of the physical shop, before it gets sorted and structured into the warehouse for the customers to buy. In a DWH you basically do the same, just with data. As you see in the DWH architecture below, the offloading area in the back of the store is your stage area where you store the source data from your operational systems or external data.

A traditional Data Warehouse architecture by Wikipedia

The physical warehouse where the customers buying the articles is in a DWH normally the so-called Data Mart. The data processed between each layer seen in the architecture above is called ETL (Extract Transform Load)t This is not to confuse with ELT (Extract Load Transform)) which is the common mythology data lakes (more in my recent post). In a DWH you always transform to get data as clean and structured as possible.

Origin: Data Engineering, the future of Data Warehousing?
References: Why having a Data Warehouse