How workloads map to user job functions

1. How workloads map to user job functions

Previously, we covered the basic building blocks of Fabric, including Fabric items. In this video, we'll look at why and how these Fabric items are organized into workloads, also known as experiences.

2. Distinguishing workloads from items

First, let's distinguish workloads from items. A workload is a broad collection of capabilities designed for a particular kind of data work. Data Engineering, Data Warehouse, and Power BI are examples of workloads. An item is an asset we create and manage inside a workspace. Lakehouses, pipelines, and warehouses are examples of items. The workload helps us find capabilities related to an area of work. The item is what we use to complete that work.

3. Moving and transforming data

Items are organized into workloads so users can quickly find the tools they need for the job they're doing. When we need to move and transform data, we find the appropriate items in the Data Factory workload. Dataflow Gen2 connects to data sources and cleans, combines, or reshapes data before loading it into Fabric storage. A pipeline suits jobs that move data or coordinate several activities as part of a larger process.

4. Turning data into business insights

To turn data into insights the business can use, we go to the Power BI workload. A semantic model defines the data structure and reusable analytical logic. A report uses the semantic model and presents that data through interactive visualizations such as charts, tables, and filters.

5. Analyzing data in real time

The Real-Time Intelligence workload contains items for analyzing data as it arrives. For example, an Eventstream might receive readings from delivery vehicles and route those readings to a destination in Fabric.

6. Storing and preparing data

The Data Engineering workload includes items that help you store, prepare, and organize data. One example is a lakehouse. A lakehouse can store files and folders, such as CSV files, alongside structured tables. You can use Spark to process large amounts of data, or use SQL to query the tables. A lakehouse can hold both the original raw data and the cleaned tables prepared for analysis.

7. Storing and querying structured business data

To store and query structured business data, such as sales, customer, or product records, we choose the Data Warehouse workload. A warehouse item stores data in relational tables and is designed for SQL-based analysis. It organizes data into schemas, which appear in the Explorer panel in the warehouse UI. A new warehouse starts with three default schemas: dbo holds business data, while sys and INFORMATION_SCHEMA hold system information. Unlike a lakehouse, a warehouse is designed for working with structured tables rather than directly storing and organizing files and folders.

8. Creating a new item

To create an item, we open the Create page, as you did earlier, select the item we want, and give it a unique name.

9. Some items exist in more than one workload

Some items exist in multiple workloads because they apply to multiple job roles. A notebook is a good example. Notebooks are interactive environments for writing code, running it, and inspecting the results. A data engineer might use a notebook to transform a large dataset. A data scientist might use one to explore data or build a machine-learning experiment.

10. Let's practice!

Let's now put this knowledge into practice!

Create Your Free Account

By continuing, you accept our Terms of Use, our Privacy Policy and that your data is stored in the USA.