[IMPALA-5100] add run time scalar subquery check

• Type: Improvement
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Component/s: Backend, Frontend
• Labels: planner ramp-up sql-language tpc-ds
• Epic Link: Subquery Support
• Epic Color: ghx-label-6
• Assignee: Unassigned
• Reporter: Greg Rahn
• Created: 20/Mar/17 06:12
• Updated: 10/Jul/20 23:35

Description

If an scalar subquery is used with an equality predicate, it must return only one row/column to be valid. If this can not be guaranteed at parse time through a single row aggregate or limit clause, Impala fails the query like such.

select i_manufact from item where i_item_sk = (select i_item_sk from item where i_item_sk = 1);
ERROR: AnalysisException: Subquery must return a single row: (SELECT i_item_sk FROM tpcds_10000_parquet.item WHERE i_item_sk = 1)

Impala should allow these to run successfully by adding a run time assert in these cases if the SQ returns > 1 row when it should not.

This impacts TPC-DS query6, query54, query58.

Issue Links

• relates to: IMPALA-1285 Subquery with group by and where could realize the result has one row.

Sub-Tasks

• Add run time scalar subquery check for uncorrelated subqueries: Resolved
• Add run time scalar subquery check for correlated subqueries: Open
• Support uncorrelated subqueries in the select list: Resolved
• Support correlated subqueries in the select list: Open