Skip to content

Apache Spark chef cookbook

License

Notifications You must be signed in to change notification settings

logicalclocks/spark-chef

Repository files navigation

Apache Spark Chef cookbook

Install Spark standalone

Install Spark yarn

References

set "spark.yarn.jars" $ Cd $SPARK_HOME $ hadoop fs mkdir spark-2.0.0-bin-hadoop $hadoop fs -copyFromLocal jars/* spark-2.0.0-bin-hadoop $ echo "spark.yarn.jars=hdfs:///nameservice1/user//spark-2.0.0-bin-hadoop/*" >> conf/spark-defaults.conf

If you do have access to the local directories of all the nodes in your cluster you can copy the archive or spark jars to the local directory of each of the data nodes using rsync or scp. Just update the URLs from hdfs:/ to local: