D-Net
/
dnet-hadoop
14
5
Fork 11
Code Issues 14 Pull Requests 12 Projects 2 Releases 2 Wiki Activity

merging with branch beta

This commit is contained in:
Miriam Baglioni 2024-07-23 12:50:12 +02:00
parent 9573bf576d d27e9ea50f
commit 62649dc5c4
2 changed files with 30 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

30
dhp-workflows/dhp-graph-mapper/src/main/resources/eu/dnetlib/dhp/oa/graph/raw_all/oozie_app/workflow.xml
View File

@ -368,6 +368,32 @@
<arg>--mdLayout</arg><arg>store</arg> <arg>--mdLayout</arg><arg>store</arg>
<arg>--mdInterpretation</arg><arg>cleaned</arg> <arg>--mdInterpretation</arg><arg>cleaned</arg>
</spark> </spark>
<ok to="ImportODF_hdfs_invisible"/>
<error to="Kill"/>
</action>
<action name="ImportODF_hdfs_invisible">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>ImportODF_hdfs_invisible</name>
<class>eu.dnetlib.dhp.oa.graph.raw.MigrateHdfsMdstoresApplication</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory ${sparkExecutorMemory}
--executor-cores ${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts>
<arg>--hdfsPath</arg><arg>${contentPath}/odf_mdstore_hdfs_invisible</arg>
<arg>--mdstoreManagerUrl</arg><arg>${mdstoreManagerUrl}</arg>
<arg>--mdFormat</arg><arg>ODF</arg>
<arg>--mdLayout</arg><arg>store</arg>
<arg>--mdInterpretation</arg><arg>intersection</arg>
</spark>
<ok to="reuse_oaf_hdfs"/> <ok to="reuse_oaf_hdfs"/>
<error to="Kill"/> <error to="Kill"/>
</action> </action>
@ -529,7 +555,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress} --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir} --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
<arg>--sourcePaths</arg><arg>${contentPath}/db_openaire,${contentPath}/db_openorgs,${contentPath}/oaf_mdstore_hdfs,${contentPath}/odf_mdstore_hdfs,${contentPath}/mdstore/*/*</arg> <arg>--sourcePaths</arg><arg>${contentPath}/db_openaire,${contentPath}/db_openorgs,${contentPath}/oaf_mdstore_hdfs,${contentPath}/odf_mdstore_hdfs,${contentPath}/odf_mdstore_hdfs_invisible,${contentPath}/mdstore/*/*</arg>
<arg>--invalidPath</arg><arg>${workingDir}/invalid_records</arg> <arg>--invalidPath</arg><arg>${workingDir}/invalid_records</arg>
<arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg> <arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
</spark> </spark>
@ -553,7 +579,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress} --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir} --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
<arg>--sourcePaths</arg><arg>${contentPath}/db_openaire,${contentPath}/db_openorgs,${contentPath}/oaf_mdstore_hdfs,${contentPath}/odf_mdstore_hdfs,${contentPath}/mdstore/*/*</arg> <arg>--sourcePaths</arg><arg>${contentPath}/db_openaire,${contentPath}/db_openorgs,${contentPath}/oaf_mdstore_hdfs,${contentPath}/odf_mdstore_hdfs,${contentPath}/odf_mdstore_hdfs_invisible,${contentPath}/mdstore/*/*</arg>
<arg>--targetPath</arg><arg>${workingDir}/entities</arg> <arg>--targetPath</arg><arg>${workingDir}/entities</arg>
<arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg> <arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
<arg>--shouldHashId</arg><arg>${shouldHashId}</arg> <arg>--shouldHashId</arg><arg>${shouldHashId}</arg>

3
dhp-workflows/dhp-graph-mapper/src/main/resources/eu/dnetlib/dhp/oa/graph/sql/queryOrganizations.sql
View File

@ -25,7 +25,8 @@ SELECT
o.country || '@@@dnet:countries' AS country, o.country || '@@@dnet:countries' AS country,
array[]::text[] AS alternativenames, array[]::text[] AS alternativenames,
'sysimport:crosswalk:entityregistry@@@dnet:provenance_actions' AS provenanceaction, 'sysimport:crosswalk:entityregistry@@@dnet:provenance_actions' AS provenanceaction,
array_remove(array_agg(DISTINCT i.pid || '###' || i.issuertype || '@@@' || i.issuertype), NULL) AS pid array_remove(array_agg(DISTINCT i.pid || '###' || i.issuertype || '@@@' || i.issuertype), NULL) AS pid,
'Unknown' AS typology
FROM dsm_organizations o FROM dsm_organizations o
LEFT OUTER JOIN dsm_services d ON (d.id = o.collectedfrom) LEFT OUTER JOIN dsm_services d ON (d.id = o.collectedfrom)
LEFT OUTER JOIN dsm_organizationpids p ON (p.organization = o.id) LEFT OUTER JOIN dsm_organizationpids p ON (p.organization = o.id)