Hi everyone, Flink 1.0.2Hadoop 2.4.0 I am running Flink on Yarn by using FlinkYarnClient to launch a Flink cluster and Flink Client to submit a PackagedProgram. To keep it simple, for batch jobs I use the WordCount example and for streaming the IterateExample and IncrementalLearning ones without args. Batch job executes successfully. However, the streaming ones fail with ClassNotFoundException. For example the IncrementalLearning job throws this exception: Caused by: java.lang.RuntimeException: Could not look up the main(String[]) method from the class org.apache.flink.streaming.examples.ml.IncrementalLearningSkeleton: org/apache/flink/streaming/api/functions/source/SourceFunction at org.apache.flink.client.program.PackagedProgram.hasMainMethod(PackagedProgram.java:479) at org.apache.flink.client.program.PackagedProgram.<init>(PackagedProgram.java:216) at org.apache.flink.client.program.PackagedProgram.<init>(PackagedProgram.java:106) [..] Caused by: java.lang.NoClassDefFoundError: org/apache/flink/streaming/api/functions/source/SourceFunction org/apache/flink/streaming/api/functions/source/SourceFunction.class at java.lang.Class.getDeclaredMethods0(Native Method) at java.lang.Class.privateGetDeclaredMethods(Class.java:2615) at java.lang.Class.getMethod0(Class.java:2856) at java.lang.Class.getMethod(Class.java:1668) at org.apache.flink.client.program.PackagedProgram.hasMainMethod(PackagedProgram.java:473) ... 45 more Caused by: java.lang.ClassNotFoundException: org.apache.flink.streaming.api.functions.source.SourceFunction 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) ... 50 more The taskmanager classpath includes the following: Classpath: /srv/hadoop-2.4.0/tmp/nm-local-dir/usercache/myuser/appcache/application_1462487692793_0012/container_1462487692793_0012_01_000002/flink-dist_2.10-1.0.2.jar It could be my pom Yarn dependency which I am not so sure about if I'm using the proper version: <dependency> <groupId>org.apache.flink</groupId> <artifactId>flink-clients_2.10</artifactId> <version>1.0.2</version> </dependency> <dependency> <groupId>org.apache.flink</groupId> <artifactId>flink-yarn_2.10</artifactId> <version>1.1-SNAPSHOT</version> </dependency> Thanks for you help! Cheers, Theo |
Hi Theo, you can't mix different Flink versions in your dependencies. Please use 1.0.2 for the flink_yarn client as well or 1.1-SNAPSHOT everywhere. On Fri, May 6, 2016 at 7:02 PM, Theofilos Kakantousis <[hidden email]> wrote:
|
Hi Robert,
Thank you for the prompt reply. You're right, it was a left over from a previous build. With the fixed dependencies, I get the same error though. A have a question on job submission as well. I use the following code to submit the job: InetSocketAddress jobManagerAddress = cluster.getJobManagerAddress(); org.apache.flink.configuration.Configuration clientConf = new org.apache.flink.configuration.Configuration(); clientConf.setInteger(ConfigConstants.JOB_MANAGER_IPC_PORT_KEY, jobManagerAddress.getPort()); clientConf.setString(ConfigConstants.JOB_MANAGER_IPC_ADDRESS_KEY, jobManagerAddress.getHostName()); Client client = new Client(clientConf); PackagedProgram program = new PackagedProgram(file,args); JobSubmissionResult res = client.runDetached(program, parallelism); Is there a way to submit the job to the cluster and avoid setting the parallelism explicitly? Thanks, Theo On 2016-05-06 19:50, Robert Metzger wrote:
|
Free forum by Nabble | Edit this page |