diff --git a/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkCreateSimRels.java b/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkCreateSimRels.java index 7bc77fe2b2..1be2b9e31e 100644 --- a/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkCreateSimRels.java +++ b/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkCreateSimRels.java @@ -88,6 +88,7 @@ public class SparkCreateSimRels extends AbstractSparkAction { JavaPairRDD mapDocuments = sc .textFile(DedupUtility.createEntityPath(graphBasePath, subEntity)) + .repartition(10000) .mapToPair( (PairFunction) s -> { MapDocument d = MapDocumentUtil.asMapDocumentWithJPath(dedupConf, s);