[HIVE-1989] recognize transitivity of predicates on join keys

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 0.6.0
• Fix Version/s: 0.8.0
• Component/s: Query Processor
• Hadoop Flags: Reviewed
• Assignee: Charles Chen Assign to me
• Reporter: John Sichi
• Created: 14/Feb/11 19:59
• Updated: 02/May/13 02:29
• Resolved: 03/Sep/11 05:37

Description

Given

set hive.mapred.mode=strict;
create table invites (foo int, bar string) partitioned by (ds string);
create table invites2 (foo int, bar string) partitioned by (ds string);
select count(*) from invites join invites2 on invites.ds=invites2.ds where invites.ds='2011-01-01';

currently an error occurs:

Error in semantic analysis: No Partition Predicate Found for Alias "invites2" Table "invites2"

The optimizer should be able to infer a predicate on invites2 via transitivity. The current lack places a burden on the user to add a redundant predicate, and makes impossible (at least in strict mode) join views where both underlying tables are partitioned (the join select list has to pick one of the tables arbitrarily).

Issue Links

• is blocked by: HIVE-1342 Predicate push down get error result when sub-queries have the same alias name
• is blocked by: HIVE-2337 Predicate pushdown erroneously conservative with outer joins
• is depended upon by: HIVE-2426 Test that views with joins work properly
• relates to: HIVE-1941 support explicit view partitioning