Launching Spark on YARN
Support for running on YARN (Hadoop NextGen) was added to Spark in version 0.6.0, and improved in 0.7.0 and 0.8.0.
Building a YARN-Enabled Assembly JAR
We need a consolidated Spark JAR (which bundles all the required dependencies) to run Spark jobs on a YARN cluster.
This can be built by setting the Hadoop version and
SPARK_YARN environment variable, as follows:
SPARK_HADOOP_VERSION=2.0.5-alpha SPARK_YARN=true ./sbt/sbt assembly
The assembled JAR will be something like this:
- Building a YARN-enabled assembly (see above).
- Your application code must be packaged into a separate JAR file.
If you want to test out the YARN deployment mode, you can use the current Spark examples. A
spark-examples_2.9.3-0.8.0-incubating file can be generated by running
sbt/sbt assembly. NOTE: since the documentation you’re reading is for Spark version 0.8.0-incubating, we are assuming here that you have downloaded Spark 0.8.0-incubating or checked it out of source control. If you are using a different version of Spark, the version numbers in the jar generated by the sbt package command will obviously be different.
Most of the configs are the same for Spark on YARN as other deploys. See the Configuration page for more information on those. These are configs that are specific to SPARK on YARN.
SPARK_YARN_USER_ENV, to add environment variables to the Spark processes launched on YARN. This can be a comma separated list of environment variables, e.g.
System Properties: * ‘spark.yarn.applicationMaster.waitTries’, property to set the number of times the ApplicationMaster waits for the the spark master and then also the number of tries it waits for the Spark Context to be intialized. Default is 10.
Launching Spark on YARN
Ensure that HADOOP_CONF_DIR or YARN_CONF_DIR points to the directory which contains the (client side) configuration files for the hadoop cluster. This would be used to connect to the cluster, write to the dfs and submit jobs to the resource manager.
The command to launch the YARN Client is as follows:
SPARK_JAR=<SPARK_ASSEMBLY_JAR_FILE> ./spark-class org.apache.spark.deploy.yarn.Client \ --jar <YOUR_APP_JAR_FILE> \ --class <APP_MAIN_CLASS> \ --args <APP_MAIN_ARGUMENTS> \ --num-workers <NUMBER_OF_WORKER_MACHINES> \ --master-memory <MEMORY_FOR_MASTER> \ --worker-memory <MEMORY_PER_WORKER> \ --worker-cores <CORES_PER_WORKER> \ --queue <queue_name>
# Build the Spark assembly JAR and the Spark examples JAR $ SPARK_HADOOP_VERSION=2.0.5-alpha SPARK_YARN=true ./sbt/sbt assembly # Configure logging $ cp conf/log4j.properties.template conf/log4j.properties # Submit Spark's ApplicationMaster to YARN's ResourceManager, and instruct Spark to run the SparkPi example $ SPARK_JAR=./assembly/target/scala-2.9.3/spark-assembly-0.8.0-incubating-hadoop2.0.5-alpha.jar \ ./spark-class org.apache.spark.deploy.yarn.Client \ --jar examples/target/scala-2.9.3/spark-examples-assembly-0.8.0-incubating.jar \ --class org.apache.spark.examples.SparkPi \ --args yarn-standalone \ --num-workers 3 \ --master-memory 4g \ --worker-memory 2g \ --worker-cores 1 # Examine the output (replace $YARN_APP_ID in the following with the "application identifier" output by the previous command) # (Note: YARN_APP_LOGS_DIR is usually /tmp/logs or $HADOOP_HOME/logs/userlogs depending on the Hadoop version.) $ cat $YARN_APP_LOGS_DIR/$YARN_APP_ID/container*_000001/stdout Pi is roughly 3.13794
The above starts a YARN Client programs which periodically polls the Application Master for status updates and displays them in the console. The client will exit once your application has finished running.
- When your application instantiates a Spark context it must use a special “yarn-standalone” master url. This starts the scheduler without forcing it to connect to a cluster. A good way to handle this is to pass “yarn-standalone” as an argument to your program, as shown in the example above.
- We do not requesting container resources based on the number of cores. Thus the numbers of cores given via command line arguments cannot be guaranteed.
- The local directories used for spark will be the local directories configured for YARN (Hadoop Yarn config yarn.nodemanager.local-dirs). If the user specifies spark.local.dir, it will be ignored.