Spark zeppelin环境配置:
export JAVA_HOME=/opt/install/jdk
export SPARK_HOME=/opt/install/spark
export HADOOP_CONF_DIR=/opt/install/hadoop/etc/hodoop
zeppelin:
val users=spark.read.options(Map(“inferSchema”->“true”,“delimiter”->",",“header”->“true”)).csv("/events/users/users.csv")
集群提交:
#在spark的home下提交:Driver在集群中的某个worker上启动,启动application
bin/spark-submit --class org.apache.spark.examples.SparkPi --master spark://nodetwo:7077 --deploy-mode cluster ./examples/jars/spark-examples_2.11-2.2.0.jar 1000
##在spark的home下运行:
bin/spark-submit --class org.apache.spark.examples.SparkPi --master spark://nodetwo:7077 ./examples/jars/spark-examples_2.11-2.2.0.jar 100
#yarn的clent默认
bin/spark-submit --class org.apache.spark.examples.SparkPi --master yarn ./examples/jars/spark-examples_2.11-2.2.0.jar 1000
#yarn的集群模式
bin/spark-submit --class