在具有HighAvailability的Hadoop集群上运行Spark-submit时出现异常

Question

I am facing Exception while running Spark-submit command on Hadoop Cluster with HighAvailability.

Following command working fine on other cluster where HA is not enabled.

spark-submit  --master yarn-client   --executor-memory 4g  --executor-cores 2    --class com.domain.app.module.mainclass   target/SNAPSHOT-jar-with-dependencies.jar

Same command not working on Cluster where HA is enabled and throwing following exception.

Exception in thread "main" java.lang.AbstractMethodError: org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider.getProxy()Lorg/apache/hadoop/io/retry/FailoverProxyProvider$ProxyInfo

Please suggest me do I need to set any congurations in spark conf.

Answer 1

From the instructions on http://hadoop.apache.org/docs/r2.6.0/hadoop-project-dist/hadoop-hdfs/HDFSHighAvailabilityWithQJM.html#Configuration_details

Please check your hdfs-site.xml:

<property>
  <name>dfs.nameservices</name>
  <value>mycluster</value> <-- Choose a name for your cluster
</property>
...
<property>
  <name>dfs.client.failover.proxy.provider.mycluster</name> <-- Put cluster name here
  <value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>
</property>

You should also check that other settings mentioned on that page are correctly configured:

• dfs.ha.namenodes.[nameservice ID]
• dfs.namenode.rpc-address.[nameservice ID].[name node ID]
• dfs.namenode.http-address.[nameservice ID].[name node ID]
• dfs.namenode.shared.edits.dir
• dfs.client.failover.proxy.provider.[nameservice ID]
• dfs.ha.fencing.methods fs.defaultFS dfs.journalnode.edits.dir