Chapter 6: Data Flows in Azure Data Factory

In this chapter, we'll look at two data flow activities: the mapping data flow and the wrangling data flow. Data flow activities provide a code-free way to implement transformations on the fly as and when data is being processed.

Incremental data loading is a very common scenario wherein data from a source is incrementally loaded to a destination. There are multiple ways to implement incremental data flows. This chapter provides an implementation of incremental data loading that you can use in your environment as and when required.

The wrangling data flow provides a code-free UI that can be used to clean and transform data using Power Query. This makes it easy for non-developers to implement data transformation and cleaning and create data pipelines quickly.

In this chapter, we'll cover the following recipes:

• Implementing incremental data loading with a mapping data flow
• Implementing a wrangling data flow

  Note...

Technical requirements

For this chapter, the following are required:

• A Microsoft Azure subscription
• PowerShell 7
• Microsoft Azure PowerShell

Implementing incremental data loading with a mapping data flow

A mapping data flow provides a code-free data flow transformation environment. We use the UI to implement the ETL and process the pipeline. Spark clusters are then provisioned, and then the data flow is transformed to Spark code and executed.

In this recipe, we'll look at one of the approaches to implement incremental data loading using a mapping data flow.

Getting ready

To get started, do the following:

1. Log in to https://portal.azure.com using your Azure credentials.
2. Open a new PowerShell prompt. Execute the following command to log in to your Azure account from PowerShell:

   Connect-AzAccount

3. You will need an existing Data Factory account. If you don't have one, create one by executing the following PowerShell script: ~/azure-data-engineering-cookbook\Chapter04\3_CreatingAzureDataFactory.ps1.
4. Create an Azure storage account and upload files to the ~/Chapter06/Data folder in orders/datain...

Implementing a wrangling data flow

A wrangling data flow performs code-free data preparation at scale by integrating Power Query to prepare/transform data. The Power Query code is converted to Spark and gets executed on a Spark cluster.

In this recipe, we'll implement a wrangling data flow to read the orders.txt file, clean the data, calculate the total sales by country and customer name, and insert the data into an Azure SQL Database table.

Getting ready

To get started, do the following:

1. Log in to https://portal.azure.com using your Azure credentials.
2. Open a new PowerShell prompt. Execute the following command to log in to your Azure account from PowerShell:

   Connect-AzAccount

3. You will need an existing Data Factory account. If you don't have one, create one by executing the ~/azure-data-engineering-cookbook\Chapter04\3_CreatingAzureDataFactory.ps1 PowerShell script.
4. Create an Azure storage account and upload the files to the ~/Chapter06/Data folder...