SingleStoreDB Spark Connector: Using a Spark job to merge a large dataset fails due to a table lock