diff --git a/dump/src/main/resources/eu/dnetlib/dhp/oa/graph/dump/eosc/oozie_app/workflow.xml b/dump/src/main/resources/eu/dnetlib/dhp/oa/graph/dump/eosc/oozie_app/workflow.xml index def9f0c..077bb4f 100644 --- a/dump/src/main/resources/eu/dnetlib/dhp/oa/graph/dump/eosc/oozie_app/workflow.xml +++ b/dump/src/main/resources/eu/dnetlib/dhp/oa/graph/dump/eosc/oozie_app/workflow.xml @@ -603,7 +603,31 @@ - + + + + + yarn + cluster + Dump for the relations between organization and projects in the subset of entities relevant for EOSC + eu.dnetlib.dhp.oa.graph.dump.eosc.SparkDumpOrganizationProject + dump-${projectVersion}.jar + + --executor-memory=${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}${spark2EventLogDir} + --conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir} + + --sourcePath${sourcePath} + --outputPath${outputPath}/dump/ + + + + eu.dnetlib.dhp.oa.graph.dump.MakeTar