Use the Cloud Storage connector with Apache Spark

This tutorial show you how to run example code that uses the Cloud Storage connector with Apache Spark.

Objectives

Write a simple wordcount Spark job in Java, Scala, or Python, then run the job on a Dataproc cluster.

Costs

This tutorial uses the following billable components of Google Cloud:

Compute Engine
Dataproc
Cloud Storage

To generate a cost estimate based on your projected usage, use the pricing calculator. New Google Cloud users might be eligible for a free trial.

Before you begin

Run the steps below to prepare to run the code in this tutorial.

Set up your project. If necessary, set up a project with the Dataproc, Compute Engine, and Cloud Storage APIs enabled and the Cloud SDK installed on your local machine.
Create a Cloud Storage bucket. You need a Cloud Storage to hold tutorial data. If you do not have one ready to use, create a new bucket in your project.
1. In the Cloud Console, go to the Cloud Storage Browser page.
  Go to Browser
2. Click Create bucket.
3. On the Create a bucket page, enter your bucket information. To go to the next step, click Continue.
  - For Name your bucket, enter a name that meets the bucket naming requirements.
  - For Choose where to store your data, do the following:
    - Select a Location type option.
    - Select a Location option.
  - For Choose a default storage class for your data, select a storage class.
  - For Choose how to control access to objects, select an Access control option.
  - For Advanced settings (optional), specify an encryption method, a retention policy, or bucket labels.
4. Click Create.
Set local environment variables. Set environment variables on your local machine. Set your Google Cloud project-id and the name of the Cloud Storage bucket you will use for this tutorial. Also provide the name and region of an existing or new Dataproc cluster. You can create a cluster to use in this tutorial in the next step.
```
PROJECT=project-id
```
```
BUCKET_NAME=bucket-name
```
```
CLUSTER=cluster-name
```
```
REGION=cluster-region Example: "us-central1"
```
Create a Dataproc cluster. Run the command, below, to create a single-node Dataproc cluster in the specified Compute Engine zone.
```
gcloud dataproc clusters create ${CLUSTER} \
    --project=${PROJECT} \
    --region=${REGION} \
    --single-node
```
The above command installs the default cluster image version. You can use the --image-version flag to select an image version for your cluster. Each image version installs specific versions of Spark and Scala library components. If you prepare the Spark wordcount job in Java or Scala, you will reference the Spark and Scala versions installed on your cluster when you prepare the job package.
Copy public data to your Cloud Storage bucket. Copy a public data Shakespeare text snippet into the input folder of your Cloud Storage bucket:
```
gsutil cp gs://pub/shakespeare/rose.txt \
    gs://${BUCKET_NAME}/input/rose.txt
```
Set up a Java (Apache Maven), Scala (SBT), or Python development environment.
Use Cloud Shell. Cloud Shell includes tools used in this tutorial, including Apache Maven, Python, and the gcloud and gsutil Cloud SDK command-line tools.

Prepare the Spark wordcount job

Select a tab, below, to follow the steps to prepare a job package or file to submit to your cluster. You can prepare one of the following job types;

Spark job in Java using Apache Maven to build a JAR package
Spark job in Scala using SBT to build a JAR package
Spark job in Python (PySpark)

Java

Copy pom.xml file to your local machine. The following pom.xml file specifies Scala and Spark library dependencies, which are given a provided scope to indicate that the Dataproc cluster will provide these libraries at runtime. The pom.xml file does not specify a Cloud Storage dependency because the connector implements the standard HDFS interface. When a Spark job accesses Cloud Storage cluster files (files with URIs that start with gs://), the system automatically uses the Cloud Storage connector to access the files in Cloud Storage

Check your cluster image version. Replace the version placeholders in the file to show the Spark and Scala library versions used by your cluster's image version. Note that the spark-core_ artifact number is the Scala major.minor version number.

<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
  <modelVersion>4.0.0</modelVersion>

  <groupId>dataproc.codelab</groupId>
  <artifactId>word-count</artifactId>
  <version>1.0</version>

  <properties>
    <maven.compiler.source>1.8</maven.compiler.source>
    <maven.compiler.target>1.8</maven.compiler.target>
  </properties>

  <dependencies>
    <dependency>
      <groupId>org.scala-lang</groupId>
      <artifactId>scala-library</artifactId>
      <version>Scala version, for example, 2.11.8</version>
      <scope>provided</scope>
    </dependency>
    <dependency>
      <groupId>org.apache.spark</groupId>
      <artifactId>spark-core_Scala major.minor.version, for example, 2.11</artifactId>
      <version>Spark version, for example, 2.3.1</version>
      <scope>provided</scope>
    </dependency>
  </dependencies>
</project>

Copy the WordCount.java code listed, below, to your local machine.

Create a set of directories with the path src/main/java/dataproc/codelab:
```
mkdir -p src/main/java/dataproc/codelab
```
Copy WordCount.java to your local machine into src/main/java/dataproc/codelab:
```
cp WordCount.java src/main/java/dataproc/codelab
```

WordCount.java is a simple Spark job in Java that reads text files from Cloud Storage, performs a word count, then writes the text file results to Cloud Storage.

package dataproc.codelab;

import java.util.Arrays;
import org.apache.spark.SparkConf;
import org.apache.spark.api.java.JavaPairRDD;
import org.apache.spark.api.java.JavaRDD;
import org.apache.spark.api.java.JavaSparkContext;
import scala.Tuple2;

public class WordCount {
  public static void main(String[] args) {
    if (args.length != 2) {
      throw new IllegalArgumentException("Exactly 2 arguments are required: <inputUri> <outputUri>");
    }
    String inputPath = args[0];
    String outputPath = args[1];
    JavaSparkContext sparkContext = new JavaSparkContext(new SparkConf().setAppName("Word Count"));
    JavaRDD<String> lines = sparkContext.textFile(inputPath);
    JavaRDD<String> words = lines.flatMap(
        (String line) -> Arrays.asList(line.split(" ")).iterator()
    );
    JavaPairRDD<String, Integer> wordCounts = words.mapToPair(
        (String word) -> new Tuple2<>(word, 1)
    ).reduceByKey(
        (Integer count1, Integer count2) -> count1 + count2
    );
    wordCounts.saveAsTextFile(outputPath);
  }
}

Build the package.
```
mvn clean package
```
If the build is successful, a target/spark-with-gcs-1.0-SNAPSHOT.jar is created.

Stage the package to Cloud Storage.

gsutil cp target/word-count-1.0.jar \
    gs://${BUCKET_NAME}/java/word-count-1.0.jar

Scala

Copy build.sbt file to your local machine. The following build.sbt file specifies Scala and Spark library dependencies, which are given a provided scope to indicate that the Dataproc cluster will provide these libraries at runtime. The build.sbt file does not specify a Cloud Storage dependency because the connector implements the standard HDFS interface. When a Spark job accesses Cloud Storage cluster files (files with URIs that start with gs://), the system automatically uses the Cloud Storage connector to access the files in Cloud Storage
Check your cluster image verison. Replace the version placeholders in the file to show the Spark and Scala library versions used by your cluster's image version.
```
scalaVersion := "Scala version, for example, 2.11.8"

name := "word-count"
organization := "dataproc.codelab"
version := "1.0"

libraryDependencies ++= Seq(
  "org.scala-lang" % "scala-library" % scalaVersion.value % "provided",
  "org.apache.spark" %% "spark-core" % "Spark version, for example, 2.3.1" % "provided"
)
```

Copy word-count.scala to your local machine. This is a simple Spark job in Java that reads text files from Cloud Storage, performs a word count, then writes the text file results to Cloud Storage.

package dataproc.codelab

import org.apache.spark.SparkContext
import org.apache.spark.SparkConf

object WordCount {
  def main(args: Array[String]) {
    if (args.length != 2) {
      throw new IllegalArgumentException(
          "Exactly 2 arguments are required: <inputPath> <outputPath>")
    }

    val inputPath = args(0)
    val outputPath = args(1)

    val sc = new SparkContext(new SparkConf().setAppName("Word Count"))
    val lines = sc.textFile(inputPath)
    val words = lines.flatMap(line => line.split(" "))
    val wordCounts = words.map(word => (word, 1)).reduceByKey(_ + _)
    wordCounts.saveAsTextFile(outputPath)
  }
}

Build the package.
```
sbt clean package
```
If the build is successful, a target/scala-2.11/word-count_2.11-1.0.jar is created.

Stage the package to Cloud Storage.

gsutil cp target/scala-2.11/word-count_2.11-1.0.jar \
    gs://${BUCKET_NAME}/scala/word-count_2.11-1.0.jar

Python

Copy word-count.py to your local machine. This is a simple Spark job in Python using PySpark that reads text files from Cloud Storage, performs a word count, then writes the text file results to Cloud Storage.

#!/usr/bin/env python

import pyspark
import sys

if len(sys.argv) != 3:
  raise Exception("Exactly 2 arguments are required: <inputUri> <outputUri>")

inputUri=sys.argv[1]
outputUri=sys.argv[2]

sc = pyspark.SparkContext()
lines = sc.textFile(sys.argv[1])
words = lines.flatMap(lambda line: line.split())
wordCounts = words.map(lambda word: (word, 1)).reduceByKey(lambda count1, count2: count1 + count2)
wordCounts.saveAsTextFile(sys.argv[2])

Submit the job

Run the following gcloud command to submit the wordcount job to your Dataproc cluster.

Java

gcloud dataproc jobs submit spark \
    --cluster=${CLUSTER} \
    --class=dataproc.codelab.WordCount \
    --jars=gs://${BUCKET_NAME}/java/word-count-1.0.jar \
    --region=${REGION} \
    -- gs://${BUCKET_NAME}/input/ gs://${BUCKET_NAME}/output/

Scala

gcloud dataproc jobs submit spark \
    --cluster=${CLUSTER} \
    --class=dataproc.codelab.WordCount \
    --jars=gs://${BUCKET_NAME}/scala/word-count_2.11-1.0.jar \
    --region=${REGION} \
    -- gs://${BUCKET_NAME}/input/ gs://${BUCKET_NAME}/output/

Python

gcloud dataproc jobs submit pyspark word-count.py \
    --cluster=${CLUSTER} \
    --region=${REGION} \
    -- gs://${BUCKET_NAME}/input/ gs://${BUCKET_NAME}/output/

View the output

After the job finishes, run the following Cloud SDK gsutil command to view the wordcount output.

gsutil cat gs://${BUCKET_NAME}/output/*

The wordcount output should be similar to the following:

(a,2)
(call,1)
(What's,1)
(sweet.,1)
(we,1)
(as,1)
(name?,1)
(any,1)
(other,1)
(rose,1)
(smell,1)
(name,1)
(would,1)
(in,1)
(which,1)
(That,1)
(By,1)

Clean up

After you finish the tutorial, you can clean up the resources that you created so that they stop using quota and incurring charges. The following sections describe how to delete or turn off these resources.

Deleting the project

The easiest way to eliminate billing is to delete the project that you created for the tutorial.

To delete the project:

In the Cloud Console, go to the Manage resources page.
Go to Manage resources
In the project list, select the project that you want to delete, and then click Delete.
In the dialog, type the project ID, and then click Shut down to delete the project.

Deleting the Dataproc cluster

Instead of deleting your project, you may wish to only delete your cluster within the project.

Deleting the Cloud Storage bucket

Cloud Console

In the Cloud Console, go to the Cloud Storage Browser page.
Go to Browser
Click the checkbox for the bucket that you want to delete.
To delete the bucket, click Delete, and then follow the instructions.

Command line

gsutil rb BUCKET_NAME

What's next

See Spark job tuning tips