1.环境下spark相关作业的目录结构
a)根目录: /opt/maintain
b)二级目录包含 jars 、scripts 、logs
c)三级目录: 项目名称,也就是项目的英文名称或简称
d)文件: 不同二级目录下三级目录下的文件含义
jars: 提交的jar包,样例:xxx-dservice-1.0.0.jar
scripts: 提交脚本,样例:spark_submit_model.sh
logs: shell脚本待输出日志存储地
2.提交模版shell
#!/bin/bash -x
## init param by your project
param1=""
param2=""
if [ $# == 1 ] ; then
param1=$1
fi
if [ $# == 2 ] ; then
param1=$1
param2=$2
fi
echo "param1:${param1}"
echo "param2:${param2}"
#basic env
JAVA_HOME=/usr/java/jdk1.8.0_181
HADOOP_CONF_DIR=/etc/hadoop/conf
SPARK_HOME=/opt/software/spark/spark
KERBEROS_USER=hdfs@HADOOP.COM
KERBEROS_USER_KEYTAB=/etc/kerberos/hdfs.keytab
#project env
PROJECT_MAIN_CLASS=com.xxx.xxx.utils.ReadPhoenixTestDemo
PROJECT_JAR="/opt/maintain/jars/sparkModel/xxx-dservice-1.0.0.