oozie创建工作流，手动配置和使用H

# 注意变量名不要包含特殊字符，否则在 spark 中会出现无法解析变量名的问题
# oozie.wf.application.path的路径必须在hdfs上，因为整个集群要访问

nameNode=hdfs://txz-data0:9820
resourceManager=txz-data0:8032
oozie.use.system.libpath=true
oozie.libpath=${nameNode}/share/lib/spark2/jars/,${nameNode}/share/lib/spark2/python/lib/,${nameNode}/share/lib/spark2/hive-site.xml
oozie.wf.application.path=${nameNode}/workflow/data-factory/download_report_voice_and_upload/Workflow
oozie.action.sharelib.for.spark=spark2

archive=${nameNode}/envs/py3.tar.gz#py

# 如果 dryrun 为 true，表示只是测试当前的 workflow，并不具体记录相应 job
dryrun=false

sparkMaster=yarn-cluster
sparkMode=cluster
scriptRoot=/workflow/data-factory/download_report_voice_and_upload/Python
sparkScriptBasename=download_parquet_from_data0_upload_online.py
sparkScript=${scriptRoot}/${sparkScriptBasename}
pysparkPath=py/py3/bin/python3

<!--
    这是为oozie的workflow提供参数，里面用到的变量默认来自job.properties文件
-->

<workflow-app xmlns='uri:oozie:workflow:1.0' name='download_parquet_from_data0_upload_online'>

    <global>
        <resource-manager>${resourceManager}</resource-manager>
        <name-node>${nameNode}</name-node>
    </global>

    <start to='spark-node' />

    <action name='spark-node'>
        <spark xmlns="uri:oozie:spark-action:1.0">
            <master>${sparkMaster}</master>
            <mode>${sparkMode}</mode>
            <name>report_voice_download_pyspark</name>
            <jar>${sparkScriptBasename}</jar>
            <spark-opts>
                --conf spark.yarn.appMasterEnv.PYSPARK_PYTHON=${pysparkPath}
            </spark-opts>
            <file>${sparkScript}#${sparkScriptBasename}</file>
            <archive>${archive}</archive>
        </spark>

        <ok to="end" />
        <error to="fail" />
    </action>

    <kill name="fail">
        <message>
            Workflow failed, error
            message[${wf:errorMessage(wf:lastErrorNode())}]
        </message>
    </kill>
    <end name='end' />
</workflow-app>