Aleksandr Efimov has posted comments on this change. ( http://gerrit.cloudera.org:8080/25001 )
Change subject: IMPALA-15059: Use HBO row counts in Calcite scans ...................................................................... Patch Set 1: (3 comments) http://gerrit.cloudera.org:8080/#/c/25001/1/java/calcite-planner/src/main/java/org/apache/impala/calcite/schema/CalciteTable.java File java/calcite-planner/src/main/java/org/apache/impala/calcite/schema/CalciteTable.java: http://gerrit.cloudera.org:8080/#/c/25001/1/java/calcite-planner/src/main/java/org/apache/impala/calcite/schema/CalciteTable.java@96 PS1, Line 96: private boolean hboStatsLoaded_; > Question: I didn't look at the hbo code closely, but are there cases where HdfsScanNode checks HBO when enabled and a key is available. PS2 stores the unfiltered lookup, including misses, in a final field at table initialization. http://gerrit.cloudera.org:8080/#/c/25001/1/java/calcite-planner/src/main/java/org/apache/impala/calcite/schema/CalciteTable.java@281 PS1, Line 281: for (FeFsPartition partition : fsTable.loadAllPartitions()) { > I'm a little concerned about duplication of running this for loop. Shared the input-row and scan-stat helpers in PS2. Physical scans still use their own file totals for pruning, sampling and simple limits; Calcite totals aren't reused yet. http://gerrit.cloudera.org:8080/#/c/25001/1/java/calcite-planner/src/main/java/org/apache/impala/calcite/schema/ImpalaRelMdRowCount.java File java/calcite-planner/src/main/java/org/apache/impala/calcite/schema/ImpalaRelMdRowCount.java: http://gerrit.cloudera.org:8080/#/c/25001/1/java/calcite-planner/src/main/java/org/apache/impala/calcite/schema/ImpalaRelMdRowCount.java@56 PS1, Line 56: public Double getRowCount(TableScan scan, RelMetadataQuery mq) { > Do we need this method? Can table.getHboRowCount() be called from table.ge Removed in PS2. CalciteTable.getRowCount() now uses HBO with catalog fallback; NULL selectivity and NDV keep using catalog row counts. -- To view, visit http://gerrit.cloudera.org:8080/25001 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I0f530098d34869cda5ff36a06576bb4854ad7f4f Gerrit-Change-Number: 25001 Gerrit-PatchSet: 1 Gerrit-Owner: Aleksandr Efimov <[email protected]> Gerrit-Reviewer: Aleksandr Efimov <[email protected]> Gerrit-Reviewer: Impala Public Jenkins <[email protected]> Gerrit-Reviewer: Quanlong Huang <[email protected]> Gerrit-Reviewer: Steve Carlin <[email protected]> Gerrit-Comment-Date: Mon, 05 Oct 2026 19:49:13 +0000 Gerrit-HasComments: Yes
