Spark ClassNotFoundException when run on yarn-cluster - scala

Question

my code:
import org.apache.spark.{SparkConf, SparkContext}
object Run extends App {
val conf = new SparkConf().setMaster("yarn-cluster").setAppName("t666")
sc.addJar("hdfs://10.1.11.99:8020/user/spark/share/scalaj-http_2.10-2.3.0.jar")
val sc = new SparkContext(conf)
val b = scalaj.http.Base64.encodeString("刘")
val a = Array[String](b)
sc.parallelize(a).saveAsTextFile("hdfs://10.1.11.99:8020/testdata/t2/")
}
and my submit commend is:
spark-submit --master yarn-cluster --class start.Run run.jar
the log on yarn show:
16/11/04 13:50:01 INFO cluster.YarnClusterScheduler: YarnClusterScheduler.postStartHook done
16/11/04 13:50:01 INFO spark.SparkContext: Added JAR hdfs://10.1.11.99:8020/user/spark/share/scalaj-http_2.10-2.3.0.jar at hdfs://10.1.11.99:8020/user/spark/share/scalaj-http_2.10-2.3.0.jar with timestamp 1478238601256
16/11/04 13:50:01 INFO cluster.YarnSchedulerBackend$YarnSchedulerEndpoint: ApplicationMaster registered as NettyRpcEndpointRef(spark://YarnAM#192.168.3.49:53976)
16/11/04 13:50:01 ERROR yarn.ApplicationMaster: User class threw exception: java.lang.NoClassDefFoundError: scalaj/http/Base64
java.lang.NoClassDefFoundError: scalaj/http/Base64
at start.Run$delayedInit$body.apply(Run.scala:31)
at scala.Function0$class.apply$mcV$sp(Function0.scala:40)
at scala.runtime.AbstractFunction0.apply$mcV$sp(AbstractFunction0.scala:12)
at scala.App$$anonfun$main$1.apply(App.scala:71)
at scala.App$$anonfun$main$1.apply(App.scala:71)
at scala.collection.immutable.List.foreach(List.scala:318)
at scala.collection.generic.TraversableForwarder$class.foreach(TraversableForwarder.scala:32)
at scala.App$class.main(App.scala:71)
at start.Run$.main(Run.scala:9)
at start.Run.main(Run.scala)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:606)
at org.apache.spark.deploy.yarn.ApplicationMaster$$anon$2.run(ApplicationMaster.scala:542)
Caused by: java.lang.ClassNotFoundException: scalaj.http.Base64
at java.net.URLClassLoader$1.run(URLClassLoader.java:366)
at java.net.URLClassLoader$1.run(URLClassLoader.java:355)
at java.security.AccessController.doPrivileged(Native Method)
at java.net.URLClassLoader.findClass(URLClassLoader.java:354)
at java.lang.ClassLoader.loadClass(ClassLoader.java:425)
at java.lang.ClassLoader.loadClass(ClassLoader.java:358)
... 15 more
16/11/04 13:50:01 INFO yarn.ApplicationMaster: Final app status: FAILED, exitCode: 15, (reason: User class threw exception: java.lang.NoClassDefFoundError: scalaj/http/Base64)
16/11/04 13:50:01 INFO client.RMProxy: Connecting to ResourceManager at slave3/192.168.3.48:8030
16/11/04 13:50:01 INFO yarn.YarnRMClient: Registering the ApplicationMaster
16/11/04 13:50:01 INFO yarn.ApplicationMaster: Started progress reporter thread with (heartbeat : 3000, initial allocation : 200) intervals
16/11/04 13:50:01 INFO spark.SparkContext: Invoking stop() from shutdown hook
the 2nd line show:
INFO spark.SparkContext: Added JAR hdfs://10.1.11.99:8020/user/spark/share/scalaj-http_2.10-2.3.0.jar at hdfs://10.1.11.99:8020/user/spark/share/scalaj-http_2.10-2.3.0.jar with timestamp 1478238601256
it seems already add the jar file into my classpath,but this exception i can't explain.
anyone's answer will be help me a lot!

Answer 1

I believe SparkContext.addJar only adds the JAR to the classpath of the workers, and not the driver. Try adding the JAR using the --jars option in the spark-submit command:
spark-submit --master yarn \
--deploy-mode cluster \
--jars hdfs://10.1.11.99:8020/user/spark/share/scalaj-http_2.10-2.3.0.jar \
--class start.Run run.jar