Github user yhuai commented on a diff in the pull request: https://github.com/apache/spark/pull/4563#discussion_r24621161 --- Diff: sql/hive/src/main/scala/org/apache/spark/sql/hive/HiveMetastoreCatalog.scala --- @@ -381,12 +391,51 @@ private[hive] class HiveMetastoreCatalog(hive: HiveContext) extends Catalog with } /** + * When scanning or writing to non-partitioned Metastore Parquet tables, convert them to Parquet + * data source relations for better performance. + */ + object ParquetConversions extends Rule[LogicalPlan] { + override def apply(plan: LogicalPlan): LogicalPlan = { + // Collects all `MetastoreRelation`s which should be replaced + val toBeReplaced = plan.collect { + case InsertIntoTable(relation: MetastoreRelation, _, _, _) + // Inserting into partitioned table is not supported in Parquet data source (yet). + if !relation.hiveQlTable.isPartitioned && + hive.convertMetastoreParquet && + hive.conf.parquetUseDataSourceApi && + relation.tableDesc.getSerdeClassName.toLowerCase.contains("parquet") => + relation + + case p @ PhysicalOperation(_, _, relation: MetastoreRelation) --- End diff -- Add a comment to explain when this one will be triggered?
--- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. --- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org