Azure Databricks & Spark For Data Engineers (PySpark / SQL)

Major updates to the course since the launch

March 2023



– New sections 6 and 7 added. Section 8 Updated. These changes reflect the latest Databricks recommendations around accessing Azure Data Lake. Also, this provides a better solution for students to complete the course project using Azure Student Subscription or Corporate Subscriptions with limited access to Azure Active Directory.

December 2022



– Sections 3, 4 & 5 updated to reflect recent UI changes to Azure Databricks and included lessons on additional functionality included by Databricks recently to Databricks clusters.

Welcome!

I look forward to helping you learn one of the in-demand data engineering tools in the cloud, Azure Databricks! This course has been taught by implementing a data engineering solution using Azure Databricks and Spark core for a real-world project of analyzing and reporting on Formula1 motor racing data.

This is like no other course in Udemy for Azure Databricks. Once you have completed the course, including all the assignments, I strongly believe that you can start a real-world data engineering project independently and be proficient in Azure Databricks. I have also included lessons on Azure Data Lake Storage Gen2, Azure Data Factory, and PowerBI. The course’s primary focus is Azure Databricks and Spark core, but it also covers the relevant concepts and connectivity to the other technologies mentioned. Please note that the course doesn’t cover other aspects of Spark, such as Spark Streaming and Spark ML. Also, the course has been taught using PySpark and Spark SQL; It doesn’t cover Scala or Java.

The course follows a logical progression of real-world project implementation, with technical concepts being explained and the Databricks notebooks being built simultaneously. Even though this course is not specifically designed to teach you the skills required for passing the Azure Data Engineer Associate Certification Exam DP203, it can greatly help you get most of the necessary skills required for the exam.

I value your time as much as I do mine. So, I have designed this course to be fast-paced and to the point. Also, the course is taught in simple English and has no jargon. I start the course from the basics, and you will be proficient in the technologies used by the end of the course.

Currently, the course teaches you the following

Azure Databricks

• Building a solution architecture for a data engineering solution using Azure Databricks, Azure Data Lake Gen2, Azure Data Factory, and Power BI
• Creating and using Azure Databricks service and the architecture of Databricks within Azure
• Working with Databricks notebooks as well as using Databricks utilities, magic commands, etc
• Passing parameters between notebooks as well as creating notebook workflows
• Creating, configuring, and monitoring Databricks clusters, cluster pools, and jobs
• Mounting Azure Storage in Databricks using secrets stored in Azure Key Vault
• Working with Databricks Tables, Databricks File System (DBFS), etc
• Using Delta Lake to implement a solution using Lakehouse architecture
• Creating dashboards to visualize the outputs
• Connecting to the Azure Databricks tables from PowerBI

Spark (Only PySpark and SQL)

• Spark architecture, Data Sources API, and Dataframe API
• PySpark – Ingestion of CSV, simple, and complex JSON files into the data lake as parquet files/ tables.
• PySpark – Transformations such as Filter, Join, Simple Aggregations, GroupBy, Window functions, etc.
• PySpark – Creating local and temporary views
• Spark SQL – Creating databases, tables, and views
• Spark SQL – Transformations such as Filter, Join, Simple Aggregations, GroupBy, Window functions, etc.
• Spark SQL – Creating local and temporary views
• Implementing full refresh and incremental load patterns using partitions

Delta Lake

• The emergence of Data Lakehouse architecture and the role of delta lake.
• Read, Write, Update, Delete, and Merge to delta lake using PySpark and SQL.
• History, Time Travel, and Vacuum
• Converting Parquet files to Delta files
• Implementing incremental load pattern using delta lake

Azure Data Factory

• Creating pipelines to execute Databricks notebooks
• Designing robust pipelines to deal with unexpected scenarios, such as missing files
• Creating dependencies between activities as well as pipelines
• Scheduling the pipelines using data factory triggers to execute at regular intervals
• Monitor the triggers/ pipelines to check for errors/ outputs.