oozie调用sub-workflow串联工作流

注意

1. 这里调度配置文件需要全部上传到hdfs，这个路径通过oozieAppsRoot指定
2. sql文件需要全部上传的hdfs,这个路径通过oozieScriptsRoot指定

配置文件

job.properties

nameNode=hdfs://node1:8020
jobTracker=node1:8032
thriftAddress=thrift://node1:9083

queueName=default

hiveSiteXml=${nameNode}/user/hive/conf/hive-site.xml

moduleName=dataclean
projectName=pv

configRoot=\({nameNode}/user/apps/\){projectName}/dataclean-script

oozieAppsRoot=\({nameNode}/user/apps/dataclean/\){projectName}

oozie.use.system.libpath=true

oozie.wf.application.path=${oozieAppsRoot}

oozieScriptsRoot=${configRoot}/dm1

projectCostScriptName=BUILD_DM1_PV_PROJECT_COST
projectDeviceScriptName=BUILD_DM1_PV_PROJECT_DEVICE

workflow.xml

\({oozieAppsRoot}/\){projectCostScriptName}

\({oozieAppsRoot}/\){projectDeviceScriptName}

Hive failed, error message[${wf:errorMessage(wf:lastErrorNode())}]

BUILD_DM1_PV_PROJECT_COST/job.properties

oozie.use.system.libpath=true

oozie.wf.application.path=\({oozieAppsRoot}/\){projectCostScriptName}

BUILD_DM1_PV_PROJECT_COST/workflow.xml

\({jobTracker}</job-tracker> <name-node>\){nameNode}
\({hiveSiteXml}</job-xml> <configuration> <property> <name>mapred.job.queue.name</name> <value>\){queueName}


hive.metastore.uris
\({thriftAddress}</value> </property> </configuration> <script>\){oozieScriptsRoot}/${projectCostScriptName}.sql

Hive failed, error message[${wf:errorMessage(wf:lastErrorNode())}]

BUILD_DM1_PV_PROJECT_DEVICE/job.properties

oozie.use.system.libpath=true

oozie.wf.application.path=\({oozieAppsRoot}/\){projectDeviceScriptName}

BUILD_DM1_PV_PROJECT_DEVICE/workflow.xml

\({jobTracker}</job-tracker> <name-node>\){nameNode}
\({hiveSiteXml}</job-xml> <configuration> <property> <name>mapred.job.queue.name</name> <value>\){queueName}


hive.metastore.uris
\({thriftAddress}</value> </property> </configuration> <script>\){oozieScriptsRoot}/${projectDeviceScriptName}.sql

Hive failed, error message[${wf:errorMessage(wf:lastErrorNode())}]