Understanding Microsoft Fabric Warehouse

Marx Fabric Warehouse looks and feels like a traditional database and in many ways it is close to that experience but under the hood it works very differently. That difference unlocks powerful features like time travel and table cloning. In this video I will give you a clear introduction to this data store with easy theory slides and practical demos in fabric. Stay tuned for more. Welcome to the video. My name is Alexian and I'm a data architect and a Microsoft data platform MVP. And on this channel I cover most fabric related topics. Today we are going to learn about the warehouse that is one of the main data stores in fabric and stores its data to one lake under the hood. But now without further ado let's get started. Let's talk about the warehouse in Marx fabric. On the side we can see how does the warehouse look in the fabric UI. Out of all the data stores in fabric the warehouse is the one that feels most like a traditional database. And many of the things that works in the traditional databases work here as well. But there are also many differences since the technology behind the warehouse is a lot different from a traditional database. Actually under the hood warehouse is storing data in Delta Park format to one lake just like a lakehouse that we covered in the previous episodes. We will do a full sidebyside comparison of fabric data stores in a later episode. But here is the main things to remember for now. The warehouse is built for structured tabular data. That means unlike the lakehouse, you can drop in just any file format. It's not meant for raw and unstructured data like images or nested JSON files. Instead, data here is stored cleanly in relational tables. Because of that, warehouse is usually used in silver and gold layers of a medallion architecture. You typically bring in cleaned and transformed data to warehouse and not raw data straight from the source. If you have used SQL Server or any relational database before, the warehouse in fabric feels right at home. But now it is packaged into fabric and packed by an open modern storage format. Now let's talk about what lives inside a warehouse. You will find the usual database objects like schemas which act like folders that group your tables and other objects together. Then tables where the actual data is stored. And then views which are just saved SQL queries that behave like virtual tables.