Qifan Chen has posted comments on this change. ( http://gerrit.cloudera.org:8080/17860 )
Change subject: IMPALA-9873: Avoid materilization of columns for filtered out rows in Parquet table. ...................................................................... Patch Set 7: (2 comments) http://gerrit.cloudera.org:8080/#/c/17860/3/be/src/exec/parquet/hdfs-parquet-scanner.cc File be/src/exec/parquet/hdfs-parquet-scanner.cc: http://gerrit.cloudera.org:8080/#/c/17860/3/be/src/exec/parquet/hdfs-parquet-scanner.cc@253 PS3, Line 253: conjuncts.reserve(scan_node_->conjuncts().size() + : scan_node_->filter_exprs().size()); : conjuncts.insert(std::end(conjuncts), std::begin(scan_node_->conjuncts()), > I tested this for Min/Max filters at row level and this code handles it as Great! May add a test for the following query. Since the filter is on a.wr_item_sk, the later materialization should help save on wr_reason_sk. select a.wr_reason_sk, a.wr_item_sk from web_returns a, web_returns b where a.wr_item_sk = b.wr_item_sk and b.wr_return_ship_cost < 10; http://gerrit.cloudera.org:8080/#/c/17860/3/be/src/exec/parquet/hdfs-parquet-scanner.cc@2330 PS3, Line 2330: > There is no recheck happening once the batch is formed even if they have fe Okay. That makes sense. -- To view, visit http://gerrit.cloudera.org:8080/17860 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I46406c913297d5bbbec3ccae62a83bb214ed2c60 Gerrit-Change-Number: 17860 Gerrit-PatchSet: 7 Gerrit-Owner: Amogh Margoor <[email protected]> Gerrit-Reviewer: Amogh Margoor <[email protected]> Gerrit-Reviewer: Impala Public Jenkins <[email protected]> Gerrit-Reviewer: Kurt Deschler <[email protected]> Gerrit-Reviewer: Qifan Chen <[email protected]> Gerrit-Reviewer: Zoltan Borok-Nagy <[email protected]> Gerrit-Comment-Date: Wed, 13 Oct 2021 17:53:28 +0000 Gerrit-HasComments: Yes
