/spark-essentials

The official repository for the Rock the JVM Spark Essentials with Scala course

Primary LanguageScala

The official repository for the Rock the JVM Spark Essentials with Scala course

This repository contains the code we wrote during Rock the JVM's Spark Essentials with Scala (Udemy version here) Unless explicitly mentioned, the code in this repository is exactly what was caught on camera.

How to install

  • install Docker
  • either clone the repo or download as zip
  • open with IntelliJ as an SBT project
  • in a terminal window, navigate to the folder where you downloaded this repo and run docker-compose up to build and start the PostgreSQL container - we will interact with it from Spark
  • in another terminal window, navigate to spark-cluster/
  • Linux/Mac users: build the Docker-based Spark cluster with
chmod +x build-images.sh
./build-images.sh
  • Windows users: build the Docker-based Spark cluster with
build-images.bat
  • when prompted to start the Spark cluster, go to the spark-cluster directory and run docker-compose up --scale spark-worker=3 to spin up the Spark containers with 3 worker nodes

A Note For Windows users: Adding Winutils

By default, Spark will be unable to write files using the local Spark executor. To write files, you will need to install the Windows Hadoop binaries, aka winutils. You can take the latest binary (Hadoop 3.2 as of June 2022), or use Hadoop 2.7 as a fallback.

After you download winutils.exe, create a directory anywhere (e.g. C:\\winutils), then create a bin directory under that, then place the winutils executable there.

You will also need to set the HADOOP_HOME environment variable to your directory where you added bin\winutils.exe. In the example above, that would be C:\\winutils.

An alternative to setting the environment variable is to add this line at the beginning of every Spark application we write:

System.setProperty("hadoop.home.dir","C:\\hadoop") // replace C:\\hadoop with your actual directory

How to start

Clone this repository and checkout the start tag by running the following in the repo folder:

git checkout start

How to see the final code

Udemy students: checkout the udemy branch of the repo:

git checkout udemy

Premium students: checkout the master branch:

git checkout master

How to run an intermediate state

The repository was built while recording the lectures. Prior to each lecture, I tagged each commit so you can easily go back to an earlier state of the repo!

The tags are as follows:

  • start
  • 1.1-scala-recap
  • 2.1-dataframes
  • 2.2-dataframes-basics-exercise
  • 2.4-datasources
  • 2.5-datasources-part-2
  • 2.6-columns-expressions
  • 2.7-columns-expressions-exercise
  • 2.8-aggregations
  • 2.9-joins
  • 2.10-joins-exercise
  • 3.1-common-types
  • 3.2-complex-types
  • 3.3-managing-nulls
  • 3.4-datasets
  • 3.5-datasets-part-2
  • 4.1-spark-sql-shell
  • 4.2-spark-sql
  • 4.3-spark-sql-exercises
  • 5.1-rdds
  • 5.2-rdds-part-2

And for premium students, in addition:

  • 6.1-spark-job-anatomy
  • 6.2-deploying-to-cluster
  • 7.1-taxi
  • 7.2-taxi-2
  • 7.3-taxi-3
  • 7.4-taxi-4

When you watch a lecture, you can git checkout the appropriate tag and the repo will go back to the exact code I had when I started the lecture.

For questions or suggestions

If you have changes to suggest to this repo, either

  • submit a GitHub issue
  • tell me in the course Q/A forum
  • submit a pull request!