Apache Spark java.lang.ClassNotFoundException

Question

Spark standalone cluster looks it's running without a problem :

http://i.stack.imgur.com/gF1fN.png

I followed this tutorial.

I have built a fat jar for running this JavaApp on the cluster. Before maven package:

find .

./pom.xml
./src
./src/main
./src/main/java
./src/main/java/SimpleApp.java

content of SimpleApp.java is :

 import org.apache.spark.api.java.*;
 import org.apache.spark.api.java.function.Function;
 import org.apache.spark.SparkConf;
 import org.apache.spark.SparkContext;


 public class SimpleApp {
 public static void main(String[] args) {

 SparkConf conf =  new SparkConf()
                   .setMaster("spark://10.35.23.13:7077")
                   .setAppName("My app")
                   .set("spark.executor.memory", "1g");

 JavaSparkContext   sc = new JavaSparkContext (conf);
 String logFile = "/home/ubuntu/spark-0.9.1/test_data";
 JavaRDD<String> logData = sc.textFile(logFile).cache();

 long numAs = logData.filter(new Function<String, Boolean>() {
  public Boolean call(String s) { return s.contains("a"); }
 }).count();

 System.out.println("Lines with a: " + numAs); 
 }
 }

This program only works when master is set as setMaster("local"). Otherwise I get this error

$java -cp path_to_file/simple-project-1.0-allinone.jar SimpleApp

http://i.stack.imgur.com/doRSn.png

Answer 1

There's the anonymous class (that extends Function) in SimpleApp.java file. This class is compiled to SimpleApp$1, which should be broadcast to each worker in the Spark cluster.

The simplest way for it is to add the jar explicitly to the Spark context. Add something like sparkContext.addJar("path_to_file/simple-project-1.0-allinone.jar") after JavaSparkContext creating and rebuild your jar file. Then the main Spark program (called the driver program) will automatically deliver your application code to the cluster.