325 lines
11 KiB
XML
325 lines
11 KiB
XML
|
<workflow-app name="enrichment_main" xmlns="uri:oozie:workflow:0.5">
|
||
|
|
||
|
<parameters>
|
||
|
<property>
|
||
|
<name>sourcePath</name>
|
||
|
<description>the source path</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>allowedsemrelsorcidprop</name>
|
||
|
<description>the semantic relationships allowed for propagation</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>allowedsemrelsresultproject</name>
|
||
|
<description>the allowed semantics </description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>allowedsemrelscommunitysemrel</name>
|
||
|
<description>the semantic relationships allowed for propagation</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>datasourceWhitelistForCountryPropagation</name>
|
||
|
<description>the white list</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>allowedtypes</name>
|
||
|
<description>the allowed types</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>outputPath</name>
|
||
|
<description>the output path</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>organizationtoresultcommunitymap</name>
|
||
|
<description>organization community map</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>pathMap</name>
|
||
|
<description>the json path associated to each selection field</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>blacklist</name>
|
||
|
<description>list of datasources in blacklist for the affiliation from instrepo propagation</description>
|
||
|
</property>
|
||
|
|
||
|
<property>
|
||
|
<name>hiveDbName</name>
|
||
|
<description>the target hive database name</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>hiveJdbcUrl</name>
|
||
|
<description>hive server jdbc url</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>hiveMetastoreUris</name>
|
||
|
<description>hive server metastore URIs</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>sparkDriverMemory</name>
|
||
|
<description>memory for driver process</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>sparkExecutorMemory</name>
|
||
|
<description>memory for individual executor</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>sparkExecutorCores</name>
|
||
|
<description>number of cores used by single executor</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>oozieActionShareLibForSpark2</name>
|
||
|
<description>oozie action sharelib for spark 2.*</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>spark2ExtraListeners</name>
|
||
|
<value>com.cloudera.spark.lineage.NavigatorAppListener</value>
|
||
|
<description>spark 2.* extra listeners classname</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>spark2SqlQueryExecutionListeners</name>
|
||
|
<value>com.cloudera.spark.lineage.NavigatorQueryListener</value>
|
||
|
<description>spark 2.* sql query execution listeners classname</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>spark2YarnHistoryServerAddress</name>
|
||
|
<description>spark 2.* yarn history server address</description>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>spark2EventLogDir</name>
|
||
|
<description>spark 2.* event log dir location</description>
|
||
|
</property>
|
||
|
</parameters>
|
||
|
|
||
|
<global>
|
||
|
<job-tracker>${jobTracker}</job-tracker>
|
||
|
<name-node>${nameNode}</name-node>
|
||
|
<configuration>
|
||
|
<property>
|
||
|
<name>mapreduce.job.queuename</name>
|
||
|
<value>${queueName}</value>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>oozie.launcher.mapred.job.queue.name</name>
|
||
|
<value>${oozieLauncherQueueName}</value>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>oozie.action.sharelib.for.spark</name>
|
||
|
<value>${oozieActionShareLibForSpark2}</value>
|
||
|
</property>
|
||
|
|
||
|
</configuration>
|
||
|
</global>
|
||
|
|
||
|
<start to="resumeFrom"/>
|
||
|
|
||
|
<kill name="Kill">
|
||
|
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
|
||
|
</kill>
|
||
|
|
||
|
<decision name="resumeFrom">
|
||
|
<switch>
|
||
|
<case to="bulk_tagging">${wf:conf('resumeFrom') eq 'BulkTagging'}</case>
|
||
|
<case to="affiliation_inst_repo">${wf:conf('resumeFrom') eq 'AffiliationInstitutionalRepository'}</case>
|
||
|
<case to="affiliation_semantic_relation">${wf:conf('resumeFrom') eq 'AffiliationSemanticRelation'}</case>
|
||
|
<case to="community_organization">${wf:conf('resumeFrom') eq 'CommunityOrganization'}</case>
|
||
|
<case to="result_project">${wf:conf('resumeFrom') eq 'ResultProject'}</case>
|
||
|
<case to="community_project">${wf:conf('resumeFrom') eq 'CommunityProject'}</case>
|
||
|
<case to="community_sem_rel">${wf:conf('resumeFrom') eq 'CommunitySemanticRelation'}</case>
|
||
|
<case to="country_propagation">${wf:conf('resumeFrom') eq 'CountryPropagation'}</case>
|
||
|
<default to="orcid_propagation"/>
|
||
|
</switch>
|
||
|
</decision>
|
||
|
|
||
|
|
||
|
<action name="orcid_propagation">
|
||
|
<sub-workflow>
|
||
|
<app-path>${wf:appPath()}/orcid_propagation
|
||
|
</app-path>
|
||
|
<propagate-configuration/>
|
||
|
<configuration>
|
||
|
<property>
|
||
|
<name>sourcePath</name>
|
||
|
<value>${sourcePath}</value>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>allowedsemrels</name>
|
||
|
<value>${allowedsemrelsorcidprop}</value>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>outputPath</name>
|
||
|
<value>${outputPath}</value>
|
||
|
</property>
|
||
|
</configuration>
|
||
|
</sub-workflow>
|
||
|
<ok to="bulk_tagging" />
|
||
|
<error to="Kill" />
|
||
|
</action>
|
||
|
|
||
|
<action name="bulk_tagging">
|
||
|
<sub-workflow>
|
||
|
<app-path>${wf:appPath()}/bulk_tagging
|
||
|
</app-path>
|
||
|
<propagate-configuration/>
|
||
|
<configuration>
|
||
|
<property>
|
||
|
<name>sourcePath</name>
|
||
|
<value>${outputPath}</value>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>baseURL</name>
|
||
|
<value>${baseURL}</value>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>pathMap</name>
|
||
|
<value>${pathMap}</value>
|
||
|
</property>
|
||
|
</configuration>
|
||
|
</sub-workflow>
|
||
|
<ok to="affiliation_inst_repo" />
|
||
|
<error to="Kill" />
|
||
|
</action>
|
||
|
|
||
|
<action name="affiliation_inst_repo">
|
||
|
<sub-workflow>
|
||
|
<app-path>${wf:appPath()}/affiliation_inst_repo
|
||
|
</app-path>
|
||
|
<propagate-configuration/>
|
||
|
<configuration>
|
||
|
<property>
|
||
|
<name>sourcePath</name>
|
||
|
<value>${outputPath}</value>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>blacklist</name>
|
||
|
<value>${blacklist}</value>
|
||
|
</property>
|
||
|
</configuration>
|
||
|
</sub-workflow>
|
||
|
<ok to="affiliation_semantic_relation" />
|
||
|
<error to="Kill" />
|
||
|
</action>
|
||
|
|
||
|
<action name="affiliation_semantic_relation">
|
||
|
<sub-workflow>
|
||
|
<app-path>${wf:appPath()}/affiliation_semantic_relation
|
||
|
</app-path>
|
||
|
<propagate-configuration/>
|
||
|
<configuration>
|
||
|
<property>
|
||
|
<name>sourcePath</name>
|
||
|
<value>${outputPath}</value>
|
||
|
</property>
|
||
|
</configuration>
|
||
|
</sub-workflow>
|
||
|
<ok to="community_organization" />
|
||
|
<error to="Kill" />
|
||
|
</action>
|
||
|
|
||
|
<action name="community_organization">
|
||
|
<sub-workflow>
|
||
|
<app-path>${wf:appPath()}/community_organization
|
||
|
</app-path>
|
||
|
<propagate-configuration/>
|
||
|
<configuration>
|
||
|
<property>
|
||
|
<name>sourcePath</name>
|
||
|
<value>${outputPath}</value>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>baseURL</name>
|
||
|
<value>${baseURL}</value>
|
||
|
</property>
|
||
|
</configuration>
|
||
|
</sub-workflow>
|
||
|
<ok to="result_project" />
|
||
|
<error to="Kill" />
|
||
|
</action>
|
||
|
|
||
|
<action name="result_project">
|
||
|
<sub-workflow>
|
||
|
<app-path>${wf:appPath()}/result_project
|
||
|
</app-path>
|
||
|
<propagate-configuration/>
|
||
|
<configuration>
|
||
|
<property>
|
||
|
<name>sourcePath</name>
|
||
|
<value>${outputPath}</value>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>allowedsemrels</name>
|
||
|
<value>${allowedsemrelsresultproject}</value>
|
||
|
</property>
|
||
|
</configuration>
|
||
|
</sub-workflow>
|
||
|
<ok to="community_project" />
|
||
|
<error to="Kill" />
|
||
|
</action>
|
||
|
|
||
|
<action name="community_project">
|
||
|
<sub-workflow>
|
||
|
<app-path>${wf:appPath()}/community_project
|
||
|
</app-path>
|
||
|
<propagate-configuration/>
|
||
|
<configuration>
|
||
|
<property>
|
||
|
<name>sourcePath</name>
|
||
|
<value>${outputPath}</value>
|
||
|
</property>
|
||
|
</configuration>
|
||
|
</sub-workflow>
|
||
|
<ok to="community_sem_rel" />
|
||
|
<error to="Kill" />
|
||
|
</action>
|
||
|
|
||
|
<action name="community_sem_rel">
|
||
|
<sub-workflow>
|
||
|
<app-path>${wf:appPath()}/community_sem_rel
|
||
|
</app-path>
|
||
|
<propagate-configuration/>
|
||
|
<configuration>
|
||
|
<property>
|
||
|
<name>sourcePath</name>
|
||
|
<value>${outputPath}</value>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>allowedsemrels</name>
|
||
|
<value>${allowedsemrelscommunitysemrel}</value>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>baseURL</name>
|
||
|
<value>${baseURL}</value>
|
||
|
</property>
|
||
|
</configuration>
|
||
|
</sub-workflow>
|
||
|
<ok to="country_propagation" />
|
||
|
<error to="Kill" />
|
||
|
</action>
|
||
|
|
||
|
<action name="country_propagation">
|
||
|
<sub-workflow>
|
||
|
<app-path>${wf:appPath()}/country_propagation
|
||
|
</app-path>
|
||
|
<propagate-configuration/>
|
||
|
<configuration>
|
||
|
<property>
|
||
|
<name>sourcePath</name>
|
||
|
<value>${outputPath}</value>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>whitelist</name>
|
||
|
<value>${datasourceWhitelistForCountryPropagation}</value>
|
||
|
</property>
|
||
|
<property>
|
||
|
<name>allowedtypes</name>
|
||
|
<value>${allowedtupes}</value>
|
||
|
</property>
|
||
|
</configuration>
|
||
|
</sub-workflow>
|
||
|
<ok to="End" />
|
||
|
<error to="Kill" />
|
||
|
</action>
|
||
|
|
||
|
<end name="End"/>
|
||
|
|
||
|
</workflow-app>
|