AWS – ページ 2 – IT研究所

AmazonS3にjavaSDKを用いて文字列を書き込む

投稿者: utsubo 投稿日: 2016-01-13 in AWS、java

ポイントは２回InputStreamを作成することです。

http://stackoverflow.com/questions/8351886/amazons3-putobject-with-inputstream-length-example

package awssample;

import java.io.ByteArrayInputStream;

import java.io.InputStream;

import java.sql.SQLException;

import com.amazonaws.ClientConfiguration;
import com.amazonaws.auth.AWSCredentials;
import com.amazonaws.auth.BasicAWSCredentials;
import com.amazonaws.services.s3.AmazonS3Client;
import com.amazonaws.services.s3.model.ObjectMetadata;
import com.amazonaws.services.s3.model.PutObjectRequest;
import com.amazonaws.util.IOUtils;

public class S3Write {

	private static String endpoint = "https://s3-ap-northeast-1.amazonaws.com";

	public static void main(String[] args) throws SQLException {

		String s = "2011-01-01,9999,1,100,1.0,-1.0,0.5";
		// 認証オブジェクトを作成
		String accessKey = "xxxxxx";
		String accessSecretKey = "xxxxxxxx";
		AWSCredentials credentials = new BasicAWSCredentials(accessKey, accessSecretKey);

		// ConfigurationでTimeout時間を30秒に設定
		ClientConfiguration clientConfiguration = new ClientConfiguration();
		clientConfiguration.setConnectionTimeout(30000);

		// AmazonS3Clientをインスタンス化
		AmazonS3Client s3 = new AmazonS3Client(credentials, clientConfiguration);
		s3.setEndpoint(endpoint);
		try {
			InputStream is = new ByteArrayInputStream(s.getBytes("UTF-8"));
			byte[] contentBytes = IOUtils.toByteArray(is);
			Long contentLength = Long.valueOf(contentBytes.length);
			System.out.println("contentLength=" + contentLength + ",s=" + s);

			ObjectMetadata metadata = new ObjectMetadata();
			metadata.setContentLength(contentLength);
											 // isではなく new ByteStreamInputStreamでもう一度さくせいしたものを渡す
			s3.putObject(new PutObjectRequest("bucketname", "path/to/file.txt",
					new ByteArrayInputStream(s.getBytes("UTF-8")), metadata));

		} catch (Exception e) {
			e.printStackTrace();
		}
	}
}

package awssample;

import java.io.ByteArrayInputStream;

import java.io.InputStream;

import java.sql.SQLException;

import com.amazonaws.ClientConfiguration;

import com.amazonaws.auth.AWSCredentials;

import com.amazonaws.auth.BasicAWSCredentials;

import com.amazonaws.services.s3.AmazonS3Client;

import com.amazonaws.services.s3.model.ObjectMetadata;

import com.amazonaws.services.s3.model.PutObjectRequest;

import com.amazonaws.util.IOUtils;

public class S3Write {

private static String endpoint = "https://s3-ap-northeast-1.amazonaws.com";

public static void main(String[] args) throws SQLException {

String s = "2011-01-01,9999,1,100,1.0,-1.0,0.5";

// 認証オブジェクトを作成

String accessKey = "xxxxxx";

String accessSecretKey = "xxxxxxxx";

AWSCredentials credentials = new BasicAWSCredentials(accessKey, accessSecretKey);

// ConfigurationでTimeout時間を30秒に設定

ClientConfiguration clientConfiguration = new ClientConfiguration();

clientConfiguration.setConnectionTimeout(30000);

// AmazonS3Clientをインスタンス化

AmazonS3Client s3 = new AmazonS3Client(credentials, clientConfiguration);

s3.setEndpoint(endpoint);

try {

InputStream is = new ByteArrayInputStream(s.getBytes("UTF-8"));

byte[] contentBytes = IOUtils.toByteArray(is);

Long contentLength = Long.valueOf(contentBytes.length);

System.out.println("contentLength=" + contentLength + ",s=" + s);

ObjectMetadata metadata = new ObjectMetadata();

metadata.setContentLength(contentLength);

// isではなく new ByteStreamInputStreamでもう一度さくせいしたものを渡す

s3.putObject(new PutObjectRequest("bucketname", "path/to/file.txt",

new ByteArrayInputStream(s.getBytes("UTF-8")), metadata));

} catch (Exception e) {

e.printStackTrace();

}

EMRでHadoopのJavaサンプル

投稿者: utsubo 投稿日: 2016-01-08 in AWS

こんな感じでディレクトリ作成します

├── bin
├── pom.xml
└── src
		└── main
				└── java
						└── emrhadoop
								├── WordCountMain.java
								├── WordCountMapper.java
								└── WordCountReducer.java

├── bin

├── pom.xml

└── src

└── main

└── java

└── emrhadoop

├── WordCountMain.java

├── WordCountMapper.java

└── WordCountReducer.java

pom.xmlを作成します

<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>jp.qri.emr</groupId>
	<artifactId>emrhive</artifactId>
	<version>1.0-SNAPSHOT</version>
	<packaging>jar</packaging>
	<properties>
		<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
	</properties>
	<repositories>
		<repository>
			<id>cloudera</id>
			<url>https://repository.cloudera.com/content/repositories/releases/</url>
		</repository>
	</repositories>
	<dependencies>
		<dependency>
			<groupId>junit</groupId>
			<artifactId>junit</artifactId>
			<version>4.12</version>
		</dependency>
		<dependency>
			<groupId>org.apache.hadoop</groupId>
			<artifactId>hadoop-core</artifactId>
			<version>1.2.1</version>
		</dependency>
	</dependencies>
	<build>
		<plugins>
						<plugin>
								<groupId>org.apache.maven.plugins</groupId>
								<artifactId>maven-dependency-plugin</artifactId>
								<configuration>
										<outputDirectory>
												${project.build.directory}
										</outputDirectory>
								</configuration>
						</plugin>
						<plugin>
								<artifactId>maven-assembly-plugin</artifactId>
								<configuration>
										<descriptorRefs>
												<descriptorRef>jar-with-dependencies</descriptorRef>
										</descriptorRefs>
								</configuration>
						</plugin>
						<plugin>
								<groupId>org.apache.maven.plugins</groupId>
								<artifactId>maven-shade-plugin</artifactId>
								<version>2.4.2</version>
								<configuration>
										<transformer
implementation="org.apache.maven.plugins.shade.resource.AppendingTransformer">
												<resource>reference.conf</resource>
										</transformer>
								</configuration>
								<executions>
										<execution>
												<phase>package</phase>
												<goals>
														<goal>shade</goal>
												</goals>
										</execution>
								</executions>
						</plugin>
				</plugins>
	</build>
</project>

<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">

<artifactId>emrhive</artifactId>

<version>1.0-SNAPSHOT</version>

<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>

</properties>

<id>cloudera</id>

<url>https://repository.cloudera.com/content/repositories/releases/</url>

</repository>

</repositories>

<groupId>junit</groupId>

<artifactId>junit</artifactId>

</dependency>

<groupId>org.apache.hadoop</groupId>

<artifactId>hadoop-core</artifactId>

</dependency>

</dependencies>

<build>

<groupId>org.apache.maven.plugins</groupId>

<artifactId>maven-dependency-plugin</artifactId>

${project.build.directory}

</outputDirectory>

</configuration>

</plugin>

<artifactId>maven-assembly-plugin</artifactId>

<descriptorRef>jar-with-dependencies</descriptorRef>

</descriptorRefs>

</configuration>

</plugin>

<groupId>org.apache.maven.plugins</groupId>

<artifactId>maven-shade-plugin</artifactId>

<transformer

implementation="org.apache.maven.plugins.shade.resource.AppendingTransformer">

<resource>reference.conf</resource>

</transformer>

</configuration>

<phase>package</phase>

<goals>

<goal>shade</goal>

</goals>

</execution>

</executions>

</plugin>

</plugins>

</build>

</project>

eclipseで読み込めるようにします

mvn eclipse:eclipse

1	mvn eclipse:eclipse

Javaファイルはこんな感じ

WordCountMain.java

package emrhadoop;

import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.input.TextInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.mapreduce.lib.output.TextOutputFormat;

public class WordCountMain {

		/**
		 * Jobを設定して実行する
		 * 
		 * @param args
		 * @throws Exception
		 */
		public static void main(String[] args) throws Exception {

				System.out.println("Masterノード start");

				// スレーブノードで実行するJobを設定する
				Job job = Job.getInstance();
				job.setJarByClass(WordCountMain.class);
				job.setJobName("wordcount");

				// Reducerへの出力キー、バリューの型を指定する
				job.setOutputKeyClass(Text.class);
				job.setOutputValueClass(IntWritable.class);

				// Mapper、Reducerのクラスを指定する
				job.setMapperClass(WordCountMapper.class);
				job.setReducerClass(WordCountReducer.class);
				// もしReducerが必要なければ、このように指定する job.setNumReduceTasks(0);

				// データを読み込み、Mapperへ渡すデータ・フォーマットを指定する
				job.setInputFormatClass(TextInputFormat.class);
				// Reducerからデータを受け取り、出力を行う際のデータ・フォーマットを指定する
				job.setOutputFormatClass(TextOutputFormat.class);

				// 引数取得
				// arg[0] は、CLIから実行した場合はメインコントローラークラス名が設定される場合もあるようだったので注意。
				String inputPath = args[0];
				System.out.println("arg 0 : " + inputPath);
				String outputPath = args[1];
				System.out.println("arg 1 : " + outputPath);

				// 入力ファイル・出力ファイルのパスを設定
				FileInputFormat.setInputPaths(job, new Path(inputPath));
				FileOutputFormat.setOutputPath(job, new Path(outputPath));

				// Job実行
				boolean result = job.waitForCompletion(true);
				System.out.println("result : " + result);

				System.out.println("Masterノード end");
		}
}

package emrhadoop;

import org.apache.hadoop.fs.Path;

import org.apache.hadoop.io.IntWritable;

import org.apache.hadoop.io.Text;

import org.apache.hadoop.mapreduce.Job;

import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;

import org.apache.hadoop.mapreduce.lib.input.TextInputFormat;

import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;

import org.apache.hadoop.mapreduce.lib.output.TextOutputFormat;

public class WordCountMain {

/**

* Jobを設定して実行する

* @param args

* @throws Exception

public static void main(String[] args) throws Exception {

System.out.println("Masterノード start");

// スレーブノードで実行するJobを設定する

Job job = Job.getInstance();

job.setJarByClass(WordCountMain.class);

job.setJobName("wordcount");

// Reducerへの出力キー、バリューの型を指定する

job.setOutputKeyClass(Text.class);

job.setOutputValueClass(IntWritable.class);

// Mapper、Reducerのクラスを指定する

job.setMapperClass(WordCountMapper.class);

job.setReducerClass(WordCountReducer.class);

// もしReducerが必要なければ、このように指定する job.setNumReduceTasks(0);

// データを読み込み、Mapperへ渡すデータ・フォーマットを指定する

job.setInputFormatClass(TextInputFormat.class);

// Reducerからデータを受け取り、出力を行う際のデータ・フォーマットを指定する

job.setOutputFormatClass(TextOutputFormat.class);

// 引数取得

// arg[0] は、CLIから実行した場合はメインコントローラークラス名が設定される場合もあるようだったので注意。

String inputPath = args[0];

System.out.println("arg 0 : " + inputPath);

String outputPath = args[1];

System.out.println("arg 1 : " + outputPath);

// 入力ファイル・出力ファイルのパスを設定

FileInputFormat.setInputPaths(job, new Path(inputPath));

FileOutputFormat.setOutputPath(job, new Path(outputPath));

// Job実行

boolean result = job.waitForCompletion(true);

System.out.println("result : " + result);

System.out.println("Masterノード end");

}

WordCountMapper.java

package emrhadoop;

import java.io.IOException;
import java.util.StringTokenizer;

import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Mapper;

/**
 * Mapper
 * 
 * 継承の際のジェネリクス指定によって、mapメソッドの型を指定出来る
 * Mapper<入力キーの型, 入力値の型, 出力キーの型, 出力値の型>
 */
public class WordCountMapper extends Mapper<LongWritable, Text, Text, IntWritable> {

		/**
		 * 初期化処理
		 */
		@Override
		public void setup(Context context) throws IOException, InterruptedException {
				System.out.println("Mapper setup");
		}

		@Override
		public void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {

				// 入力値を取り出す（1行データ)
				String line = value.toString();

				// 単語に分解する
				StringTokenizer tokenizer = new StringTokenizer(line);

				IntWritable one = new IntWritable(1);
				Text word = new Text();

				// 単語ごとに繰り返し
				while (tokenizer.hasMoreTokens()) {
						word.set(tokenizer.nextToken());

						// 1単語ごとにReducerへ値を渡す。(単語, 集計数)。ここでは単純に1単語につき1を渡しているだけだが、Mapper側で一度集計してからReducerに渡してもいい。
						context.write(word, one);
				}
		}

		/**
		 * 終了処理
		 */
		@Override
		public void cleanup(Context context) throws IOException,InterruptedException {
				System.out.println("Mapper cleanup");
		}
}

package emrhadoop;

import java.io.IOException;

import java.util.StringTokenizer;

import org.apache.hadoop.io.IntWritable;

import org.apache.hadoop.io.LongWritable;

import org.apache.hadoop.io.Text;

import org.apache.hadoop.mapreduce.Mapper;

/**

* Mapper

* 継承の際のジェネリクス指定によって、mapメソッドの型を指定出来る

* Mapper<入力キーの型, 入力値の型, 出力キーの型, 出力値の型>

public class WordCountMapper extends Mapper<LongWritable, Text, Text, IntWritable> {

/**

* 初期化処理

@Override

public void setup(Context context) throws IOException, InterruptedException {

System.out.println("Mapper setup");

}

@Override

public void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {

// 入力値を取り出す（1行データ)

String line = value.toString();

// 単語に分解する

StringTokenizer tokenizer = new StringTokenizer(line);

IntWritable one = new IntWritable(1);

Text word = new Text();

// 単語ごとに繰り返し

while (tokenizer.hasMoreTokens()) {

word.set(tokenizer.nextToken());

// 1単語ごとにReducerへ値を渡す。(単語, 集計数)。ここでは単純に1単語につき1を渡しているだけだが、Mapper側で一度集計してからReducerに渡してもいい。

context.write(word, one);

}

/**

* 終了処理

@Override

public void cleanup(Context context) throws IOException,InterruptedException {

System.out.println("Mapper cleanup");

}

WordCountReducer.java

package emrhadoop;

import java.io.IOException;

import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Reducer;

/**
 * Reducer
 * 
 * 継承の際のジェネリクス指定によって、reduceメソッドの型を指定出来る
 * Reducer<入力キーの型, 入力値の型, 出力キーの型, 出力値の型>
 */
public class WordCountReducer extends Reducer<Text, IntWritable, Text, IntWritable> {

		@Override
		public void reduce(Text key, Iterable<IntWritable> values, Context context) throws IOException, InterruptedException {

				// Mapperから渡された値を集計
				int sum = 0;
				for (IntWritable value : values) {
						sum += value.get();
				}

				// 書き込み
				context.write(key, new IntWritable(sum));
		}

}

package emrhadoop;

import java.io.IOException;

import org.apache.hadoop.io.IntWritable;

import org.apache.hadoop.io.Text;

import org.apache.hadoop.mapreduce.Reducer;

/**

* Reducer

* 継承の際のジェネリクス指定によって、reduceメソッドの型を指定出来る

* Reducer<入力キーの型, 入力値の型, 出力キーの型, 出力値の型>

public class WordCountReducer extends Reducer<Text, IntWritable, Text, IntWritable> {

@Override

public void reduce(Text key, Iterable<IntWritable> values, Context context) throws IOException, InterruptedException {

// Mapperから渡された値を集計

int sum = 0;

for (IntWritable value : values) {

sum += value.get();

}

// 書き込み

context.write(key, new IntWritable(sum));

}

Jar作成

mvn package

1	mvn package

AWS Console

まずEMRを作成します
Cleate Clusterから Go to advanced optionsへ
Hardware ConfigurationでEC2 instance typeを必要に応じ変更。m1.mediumが最安かな？
VPCに対応したのでVPC内に作成したい場合にはここで選択
その他、キーとかSecurityGroupなどを適宜設定します
　その後、作成したJarをS3へコピーしておきます
WordCount用のファイルをS3へコピーします

aws s3 cp input.txt s3://bucket/input/

1	aws s3 cp input.txt s3://bucket/input/

Stepsから起動します

StepTypeはCustomJARを選択
JAR locationに先ほどコピーしたJarファイルのS3のLocationを入力
Argumentsに

emrhadoop.WordCountMain s3n://bucket/input/input.txt s3n://bucket/output

1	emrhadoop.WordCountMain s3n://bucket/input/input.txt s3n://bucket/output

ちなみにouputディレクトリをあらかじめ作成しておくとエラーになります

AmazonEMRでSparkを動かす

投稿者: utsubo 投稿日: 2015-12-22 in AWS

結構はまってしまったのでメモ

VPC内に作成したEMRでSparkを動かしました。サンプルはいろいろなところにそこそこあるのですが、どうもきちっと動くものがなく結構苦労してしまいました。

EMR

まずはVPC内にEMRを作成します。EMRのコンソール画面を開き、CreateClusterを押します。

VPC内に作成するためには上の方にあるGo to Advanced optionから進む必要があります。

VPCとサブネットを設定し、パーミッションなどを設定しクラスターを作成します。

計算プログラム

こんな感じでディレクトリ作成します

なおscala 2.10,jdk1.8を使用しています

s3にあらかじめbucketという名のbucketとその下にoutputディレクトリを作成し、EMRから触れるようにパーミッションを設定しておきます

├── build.sbt
├── project
│　└── assembly.sbt
├── src
│　├── main
│　 │　├── java
│　 │　 ├── resources
│　 │　└── scala
│　 │　		 └── sample
│　│　				└── SparkPi.scala
│　└── test
│　		├── resources
│　		└── scala
└── target

├── build.sbt

├── project

│　└── assembly.sbt

├── src

│　├── main

│　 │　├── java

│　 │　 ├── resources

│　 │　└── scala

│　 │　 └── sample

│　│　 └── SparkPi.scala

│　└── test

│　 ├── resources

│　 └── scala

└── target

build.sbt

name := "emrscala"

version := "0.0.1"

scalaVersion := "2.10.5"

libraryDependencies ++= Seq(
	("org.apache.spark" %% "spark-sql" % "1.3.1").
		exclude("org.mortbay.jetty", "servlet-api").
		exclude("com.google.guava","guava").
		exclude("org.apache.hadoop","hadoop-yarn-api").
		exclude("commons-beanutils", "commons-beanutils-core").
		exclude("commons-beanutils", "commons-beanutils").
		exclude("commons-collections", "commons-collections").
		exclude("commons-logging", "commons-logging").
		exclude("org.spark-project.spark", "unused"). 
		exclude("com.twitter", "parquet-encoding").
		exclude("com.twitter", "parquet-column").
		exclude("com.twitter", "parquet-hadoop-bundle").
		exclude("org.datanucleus", "datanucleus-api-jdo").
		exclude("org.datanucleus", "datanucleus-core").
		exclude("org.datanucleus", "datanucleus-rdbms").
		exclude("com.esotericsoftware.minlog", "minlog"),
	("org.apache.spark" %% "spark-mllib" % "1.3.1").
		exclude("org.mortbay.jetty", "servlet-api").
		exclude("com.google.guava","guava").
		exclude("org.apache.hadoop","hadoop-yarn-api").
		exclude("commons-beanutils", "commons-beanutils-core").
		exclude("commons-beanutils", "commons-beanutils").
		exclude("commons-collections", "commons-collections").
		exclude("commons-logging", "commons-logging").
		exclude("org.spark-project.spark", "unused"). 
		exclude("com.twitter", "parquet-encoding").
		exclude("com.twitter", "parquet-column").
		exclude("com.twitter", "parquet-hadoop-bundle").
		exclude("org.datanucleus", "datanucleus-api-jdo").
		exclude("org.datanucleus", "datanucleus-core").
		exclude("org.datanucleus", "datanucleus-rdbms").
		exclude("com.esotericsoftware.minlog", "minlog"),
	("org.apache.spark" %% "spark-hive" % "1.3.1").
		exclude("org.mortbay.jetty", "servlet-api").
		exclude("com.google.guava","guava").
		exclude("org.apache.hadoop","hadoop-yarn-api").
		exclude("commons-beanutils", "commons-beanutils-core").
		exclude("commons-beanutils", "commons-beanutils").
		exclude("commons-collections", "commons-collections").
		exclude("commons-logging", "commons-logging").
		exclude("org.spark-project.spark", "unused"). 
		exclude("com.twitter", "parquet-encoding").
		exclude("com.twitter", "parquet-column").
		exclude("com.twitter", "parquet-hadoop-bundle").
		exclude("org.datanucleus", "datanucleus-api-jdo").
		exclude("org.datanucleus", "datanucleus-core").
		exclude("org.datanucleus", "datanucleus-rdbms").
		exclude("com.esotericsoftware.minlog", "minlog"),
	("org.apache.spark" %% "spark-sql" % "1.3.1").
		exclude("org.mortbay.jetty", "servlet-api").
		exclude("com.google.guava","guava").
		exclude("org.apache.hadoop","hadoop-yarn-api").
		exclude("commons-beanutils", "commons-beanutils-core").
		exclude("commons-beanutils", "commons-beanutils").
		exclude("commons-collections", "commons-collections").
		exclude("commons-logging", "commons-logging").
		exclude("org.spark-project.spark", "unused"). 
		exclude("com.twitter", "parquet-encoding").
		exclude("com.twitter", "parquet-column").
		exclude("com.twitter", "parquet-hadoop-bundle").
		exclude("org.datanucleus", "datanucleus-api-jdo").
		exclude("org.datanucleus", "datanucleus-core").
		exclude("org.datanucleus", "datanucleus-rdbms").
		exclude("com.esotericsoftware.minlog", "minlog"),
	("org.apache.spark" %% "spark-core" % "1.3.1").
		exclude("org.mortbay.jetty", "servlet-api").
		exclude("com.google.guava","guava").
		exclude("org.apache.hadoop","hadoop-yarn-api").
		exclude("commons-beanutils", "commons-beanutils-core").
		exclude("commons-beanutils", "commons-beanutils").
		exclude("commons-collections", "commons-collections").
		exclude("commons-logging", "commons-logging").
		exclude("org.spark-project.spark", "unused"). 
		exclude("com.twitter", "parquet-encoding").
		exclude("com.twitter", "parquet-column").
		exclude("com.twitter", "parquet-hadoop-bundle").
		exclude("org.datanucleus", "datanucleus-api-jdo").
		exclude("org.datanucleus", "datanucleus-core").
		exclude("org.datanucleus", "datanucleus-rdbms").
		exclude("com.esotericsoftware.minlog", "minlog")
)

name := "emrscala"

version := "0.0.1"

scalaVersion := "2.10.5"

libraryDependencies ++= Seq(

("org.apache.spark" %% "spark-sql" % "1.3.1").

exclude("org.mortbay.jetty", "servlet-api").

exclude("com.google.guava","guava").

exclude("org.apache.hadoop","hadoop-yarn-api").

exclude("commons-beanutils", "commons-beanutils-core").

exclude("commons-beanutils", "commons-beanutils").

exclude("commons-collections", "commons-collections").

exclude("commons-logging", "commons-logging").

exclude("org.spark-project.spark", "unused").

exclude("com.twitter", "parquet-encoding").

exclude("com.twitter", "parquet-column").

exclude("com.twitter", "parquet-hadoop-bundle").

exclude("org.datanucleus", "datanucleus-api-jdo").

exclude("org.datanucleus", "datanucleus-core").

exclude("org.datanucleus", "datanucleus-rdbms").

exclude("com.esotericsoftware.minlog", "minlog"),

("org.apache.spark" %% "spark-mllib" % "1.3.1").

exclude("org.mortbay.jetty", "servlet-api").

exclude("com.google.guava","guava").

exclude("org.apache.hadoop","hadoop-yarn-api").

exclude("commons-beanutils", "commons-beanutils-core").

exclude("commons-beanutils", "commons-beanutils").

exclude("commons-collections", "commons-collections").

exclude("commons-logging", "commons-logging").

exclude("org.spark-project.spark", "unused").

exclude("com.twitter", "parquet-encoding").

exclude("com.twitter", "parquet-column").

exclude("com.twitter", "parquet-hadoop-bundle").

exclude("org.datanucleus", "datanucleus-api-jdo").

exclude("org.datanucleus", "datanucleus-core").

exclude("org.datanucleus", "datanucleus-rdbms").

exclude("com.esotericsoftware.minlog", "minlog"),

("org.apache.spark" %% "spark-hive" % "1.3.1").

exclude("org.mortbay.jetty", "servlet-api").

exclude("com.google.guava","guava").

exclude("org.apache.hadoop","hadoop-yarn-api").

exclude("commons-beanutils", "commons-beanutils-core").

exclude("commons-beanutils", "commons-beanutils").

exclude("commons-collections", "commons-collections").

exclude("commons-logging", "commons-logging").

exclude("org.spark-project.spark", "unused").

exclude("com.twitter", "parquet-encoding").

exclude("com.twitter", "parquet-column").

exclude("com.twitter", "parquet-hadoop-bundle").

exclude("org.datanucleus", "datanucleus-api-jdo").

exclude("org.datanucleus", "datanucleus-core").

exclude("org.datanucleus", "datanucleus-rdbms").

exclude("com.esotericsoftware.minlog", "minlog"),

("org.apache.spark" %% "spark-sql" % "1.3.1").

exclude("org.mortbay.jetty", "servlet-api").

exclude("com.google.guava","guava").

exclude("org.apache.hadoop","hadoop-yarn-api").

exclude("commons-beanutils", "commons-beanutils-core").

exclude("commons-beanutils", "commons-beanutils").

exclude("commons-collections", "commons-collections").

exclude("commons-logging", "commons-logging").

exclude("org.spark-project.spark", "unused").

exclude("com.twitter", "parquet-encoding").

exclude("com.twitter", "parquet-column").

exclude("com.twitter", "parquet-hadoop-bundle").

exclude("org.datanucleus", "datanucleus-api-jdo").

exclude("org.datanucleus", "datanucleus-core").

exclude("org.datanucleus", "datanucleus-rdbms").

exclude("com.esotericsoftware.minlog", "minlog"),

("org.apache.spark" %% "spark-core" % "1.3.1").

exclude("org.mortbay.jetty", "servlet-api").

exclude("com.google.guava","guava").

exclude("org.apache.hadoop","hadoop-yarn-api").

exclude("commons-beanutils", "commons-beanutils-core").

exclude("commons-beanutils", "commons-beanutils").

exclude("commons-collections", "commons-collections").

exclude("commons-logging", "commons-logging").

exclude("org.spark-project.spark", "unused").

exclude("com.twitter", "parquet-encoding").

exclude("com.twitter", "parquet-column").

exclude("com.twitter", "parquet-hadoop-bundle").

exclude("org.datanucleus", "datanucleus-api-jdo").

exclude("org.datanucleus", "datanucleus-core").

exclude("org.datanucleus", "datanucleus-rdbms").

exclude("com.esotericsoftware.minlog", "minlog")

)

assembly.sbt

addSbtPlugin("com.eed3si9n" % "sbt-assembly" % "0.14.1")

addSbtPlugin("com.typesafe.sbteclipse" % "sbteclipse-plugin" % "4.0.0")

addSbtPlugin("com.eed3si9n" % "sbt-assembly" % "0.14.1")

addSbtPlugin("com.typesafe.sbteclipse" % "sbteclipse-plugin" % "4.0.0")

SparkPi.scala

package sample
import scala.math.random

import org.apache.spark.mllib.util.MLUtils
import org.apache.spark._

object SparkPi {
	def main(args: Array[String]) {
		val conf = new SparkConf().setAppName("Spark Pi").setMaster("local[2]")
		val spark = new SparkContext(conf)
		val slices=2
		val n = math.min(100000L * slices, Int.MaxValue).toInt // avoid overflow
		val count = spark.parallelize(1 until n, slices).map { i =>
			val x = random * 2 - 1
			val y = random * 2 - 1
			if (x*x + y*y < 1) 1 else 0
		}.reduce(_ + _)
		println("Pi is roughly " + 4.0 * count / n)
		
		val outputLocation = "s3n://bucket/output"
		val pi=4.0 * count / n
		val data=spark.makeRDD(Seq(pi))
		data.saveAsTextFile(outputLocation + "/pi")
	spark.stop()
	}
}

package sample

import scala.math.random

import org.apache.spark.mllib.util.MLUtils

import org.apache.spark._

object SparkPi {

def main(args: Array[String]) {

val conf = new SparkConf().setAppName("Spark Pi").setMaster("local[2]")

val spark = new SparkContext(conf)

val slices=2

val n = math.min(100000L * slices, Int.MaxValue).toInt // avoid overflow

val count = spark.parallelize(1 until n, slices).map { i =>

val x = random * 2 - 1

val y = random * 2 - 1

if (x*x + y*y < 1) 1 else 0

}.reduce(_ + _)

println("Pi is roughly " + 4.0 * count / n)

val outputLocation = "s3n://bucket/output"

val pi=4.0 * count / n

val data=spark.makeRDD(Seq(pi))

data.saveAsTextFile(outputLocation + "/pi")

spark.stop()

}

make

sbt assembly

1	sbt assembly

できたjarファイルをS3にコピーします

実行

EMRのadd StepからCustom JARを選択

JARLocationに先ほどアップしたJarを選択

Argumentに

--verbose sample.SparkPi

1	--verbose sample.SparkPi

こんな感じで実行

しばらくたつと s3://bucket/output/piいかに結果が格納されています。

AWS EMR でSparkRを使って見る

投稿者: utsubo 投稿日: 2015-12-03 in AWS、R

AWSEMRとは、SparkやらHiveやらそれら一式を簡単に使える様にしてくれている仕組みです。

ぽちぽちっとEMRでサーバを作成。

この間１０分程度

SparkRでサンプルデータを解析してみます

こちらの内容をアレンジしてみました

http://engineer.recruit-lifestyle.co.jp/techblog/2015-08-19-sparkr/

データ取得

http://stat-computing.org/dataexpo/2009/the-data.html

こちらから２００１、２、３のデータをダウンロード

$ wget http://stat-computing.org/dataexpo/2009/2001.csv.bz2

1	$ wget http://stat-computing.org/dataexpo/2009/2001.csv.bz2

unzip

$ bunzip2 2001.csv.bz2

1	$ bunzip2 2001.csv.bz2

s3にアップロード

$ aws s3 cp 2001.csv s3://samplebucket/airline/

1	$ aws s3 cp 2001.csv s3://samplebucket/airline/

同様に2002,2003も繰り返す

Hive

$ hive
hive> add jar /usr/lib/hive/lib/hive-contrib.jar;
Added [/usr/lib/hive/lib/hive-contrib.jar] to class path
Added resources: [/usr/lib/hive/lib/hive-contrib.jar]
hive> create table airline(
		> Year STRING,
		> Month STRING,
		> DayofMonth STRING,
		> DayOfWeek STRING,
		> DepTime STRING,
		> CRSDepTime STRING,
		> ArrTime STRING,
		> CRSArrTime STRING,
		> UniqueCarrier STRING,
		> FlightNum STRING,
		> TailNum STRING,
		> ActualElapsedTime STRING,
		> CRSElapsedTime STRING,
		> AirTime STRING,
		> ArrDelay STRING,
		> DepDelay STRING,
		> Origin STRING,
		> Dest STRING,
		> Distance STRING,
		> TaxiIn STRING,
		> TaxiOut STRING,
		> Cancelled STRING,
		> CancellationCode STRING,
		> Diverted STRING,
		> CarrierDelay STRING,
		> WeatherDelay STRING,
		> NASDelay STRING,
		> SecurityDelay STRING,
		> LateAircraftDelay STRING
		> )
		> ROW FORMAT DELIMITED FIELDS TERMINATED BY ',' LINES TERMINATED BY '\n'
		> LOCATION 's3://samplebucket/airline/' tblproperties ("skip.header.line.count"="1");
hive> select * from airline limit 1;
OK
2001	1	17	3	1806	1810	1931	1934	US	375	N700&#65533;&#65533;	85	84	60	-3	-4	BWI	CLT	361	5	20	0	NA	0	NA	NA	NA	NA	NA

$ hive

hive> add jar /usr/lib/hive/lib/hive-contrib.jar;

Added [/usr/lib/hive/lib/hive-contrib.jar] to class path

Added resources: [/usr/lib/hive/lib/hive-contrib.jar]

hive> create table airline(

> Year STRING,

> Month STRING,

> DayofMonth STRING,

> DayOfWeek STRING,

> DepTime STRING,

> CRSDepTime STRING,

> ArrTime STRING,

> CRSArrTime STRING,

> UniqueCarrier STRING,

> FlightNum STRING,

> TailNum STRING,

> ActualElapsedTime STRING,

> CRSElapsedTime STRING,

> AirTime STRING,

> ArrDelay STRING,

> DepDelay STRING,

> Origin STRING,

> Dest STRING,

> Distance STRING,

> TaxiIn STRING,

> TaxiOut STRING,

> Cancelled STRING,

> CancellationCode STRING,

> Diverted STRING,

> CarrierDelay STRING,

> WeatherDelay STRING,

> NASDelay STRING,

> SecurityDelay STRING,

> LateAircraftDelay STRING

> )

> ROW FORMAT DELIMITED FIELDS TERMINATED BY ',' LINES TERMINATED BY '\n'

> LOCATION 's3://samplebucket/airline/' tblproperties ("skip.header.line.count"="1");

hive> select * from airline limit 1;

2001 1 17 3 1806 1810 1931 1934 US 375 N700�� 85 84 60 -3 -4 BWI CLT 361 5 20 0 NA 0 NA NA NA NA NA

SparkR

$ sparkR
> install.packages("magrittr")
> library(magrittr)
> hiveContext <- sparkRHive.init(sc)
> airline<-sql(hiveContext,"select * from airline")
> class(airline)
[1] "DataFrame"
attr(,"package")
[1] "SparkR"
> airline %>%
+	 filter(airline$Origin == "JFK") %>%
+	 group_by(airline$Dest) %>%
+	 agg(count=n(airline$Dest)) %>%
+	 head
	Dest count																																		
1	IAH	1214
2	STL	2922
3	SNA	 805
4	MSP	1580
5	STT	1085
6	SAN	2723

$ sparkR

> install.packages("magrittr")

> library(magrittr)

> hiveContext <- sparkRHive.init(sc)

> airline<-sql(hiveContext,"select * from airline")

> class(airline)

[1] "DataFrame"

attr(,"package")

[1] "SparkR"

> airline %>%

+ filter(airline$Origin == "JFK") %>%

+ group_by(airline$Dest) %>%

+ agg(count=n(airline$Dest)) %>%

+ head

Dest count

1 IAH 1214

2 STL 2922

3 SNA 805

4 MSP 1580

5 STT 1085

6 SAN 2723

こんな感じで簡単にできました