[DRILL-3380] CTAS Auto Partitioning : We are not pruning when we use functions in the select list

• Type: Bug
• Status: Closed
• Priority: Critical
• Resolution: Fixed
• Fix Version/s: 1.1.0
• Component/s: Query Planning & Optimization
• Assignee: Steven Phillips
• Reporter: Rahul Kumar Challapalli
• Created: 26/Jun/15 00:33
• Updated: 11/Dec/15 23:55
• Resolved: 26/Jun/15 19:12

Description

git.commit.id.abbrev=5a34d81

I used the below query to create a paritioned data set

create table `lineitem` partition by (l_moddate) as select l.*, l_shipdate - extract(day from l_shipdate) + 1 l_moddate from cp.`tpch/lineitem.parquet` l;

The plan for the below query only scans one file

explain plan for select * from `lineitem` where l_moddate = date '1994-07-01';
 00-00    Screen
00-01      Project(*=[$0])
00-02        Project(*=[$0])
00-03          Scan(groupscan=[ParquetGroupScan [entries=[ReadEntryWithPath [path=/drill/testdata/ctas_auto_partition/tpch_single_partition/lineitem/0_0_31.parquet]], selectionRoot=/drill/testdata/ctas_auto_partition/tpch_single_partition/lineitem, numFiles=1, columns=[`*`]]])

However the below plan indicates a full table scan

explain plan for select count(*) from `tpch_single_partition/lineitem` where l_moddate = date '1994-07-01';
00-00    Screen
00-01      StreamAgg(group=[{}], EXPR$0=[COUNT()])
00-02        Project($f0=[0])
00-03          SelectionVectorRemover
00-04            Filter(condition=[=($0, 1994-07-01)])
00-05              Scan(groupscan=[ParquetGroupScan [entries=[ReadEntryWithPath [path=maprfs:///drill/testdata/ctas_auto_partition/tpch_single_partition/lineitem]], selectionRoot=/drill/testdata/ctas_auto_partition/tpch_single_partition/lineitem, numFiles=1, columns=[`l_moddate`]]])

Issue Links

• is duplicated by: DRILL-3379 Passing references when cloning ParquetGroupScan causes incorrect planning