Can I get a review for a small fix in pinot-spark-connector: https://github.com/apache/pinot/pull/10209 ? This addresses some memory issue on Spark executor when scanning large segments via GRPC server.