Skip to content

Enable Parallel-oblivious Hash Left Anti Semi (Not-In) Join #30

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
avamingli merged 1 commit into from
Aug 9, 2023
Merged

Enable Parallel-oblivious Hash Left Anti Semi (Not-In) Join #30

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion src/backend/optimizer/path/joinpath.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2299,7 +2299,6 @@ hash_inner_and_outer(PlannerInfo *root,
save_jointype != JOIN_UNIQUE_OUTER &&
save_jointype != JOIN_FULL &&
save_jointype != JOIN_RIGHT &&
save_jointype != JOIN_LASJ_NOTIN &&
save_jointype != JOIN_DEDUP_SEMI &&
save_jointype != JOIN_DEDUP_SEMI_REVERSE &&
outerrel->partial_pathlist != NIL &&
Expand All @@ -2319,6 +2318,7 @@ hash_inner_and_outer(PlannerInfo *root,
*/
if (innerrel->partial_pathlist != NIL &&
save_jointype != JOIN_UNIQUE_INNER &&
save_jointype != JOIN_LASJ_NOTIN &&
enable_parallel_hash)
{
cheapest_partial_inner =
Expand Down
75 changes: 75 additions & 0 deletions src/test/regress/expected/gp_parallel.out
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1590,6 +1590,81 @@ select * from t1 order by c2 asc limit 3 offset 5;

abort;
--
-- Test Parallel Hash Left Anti Semi (Not-In) Join(parallel-oblivious).
--
create table t1(c1 int, c2 int) using ao_row distributed by (c1);
create table t2(c1 int, c2 int) using ao_row distributed by (c1);
create table t3_null(c1 int, c2 int) using ao_row distributed by (c1);
set enable_parallel = on;
set gp_appendonly_insert_files = 2;
set gp_appendonly_insert_files_tuples_range = 100;
set max_parallel_workers_per_gather = 2;
insert into t1 select i, i from generate_series(1, 5000000) i;
insert into t2 select i+1, i from generate_series(1, 1200) i;
insert into t3_null select i+1, i from generate_series(1, 1200) i;
insert into t3_null values(NULL, NULL);
analyze t1;
analyze t2;
analyze t3_null;
explain(costs off) select sum(t1.c1) from t1 where c1 not in (select c1 from t2);
QUERY PLAN
---------------------------------------------------------------------------
Finalize Aggregate
-> Gather Motion 6:1 (slice1; segments: 6)
-> Partial Aggregate
-> Hash Left Anti Semi (Not-In) Join
Hash Cond: (t1.c1 = t2.c1)
-> Parallel Seq Scan on t1
-> Hash
-> Broadcast Motion 3:6 (slice2; segments: 3)
-> Seq Scan on t2
Optimizer: Postgres query optimizer
(10 rows)

select sum(t1.c1) from t1 where c1 not in (select c1 from t2);
sum
----------------
12500001778200
(1 row)

explain(costs off) select * from t1 where c1 not in (select c1 from t3_null);
QUERY PLAN
---------------------------------------------------------------
Gather Motion 6:1 (slice1; segments: 6)
-> Hash Left Anti Semi (Not-In) Join
Hash Cond: (t1.c1 = t3_null.c1)
-> Parallel Seq Scan on t1
-> Hash
-> Broadcast Motion 3:6 (slice2; segments: 3)
-> Seq Scan on t3_null
Optimizer: Postgres query optimizer
(8 rows)

select * from t1 where c1 not in (select c1 from t3_null);
c1 | c2
----+----
(0 rows)

-- non-parallel results.
set enable_parallel = off;
select sum(t1.c1) from t1 where c1 not in (select c1 from t2);
sum
----------------
12500001778200
(1 row)

select * from t1 where c1 not in (select c1 from t3_null);
c1 | c2
----+----
(0 rows)

drop table t1;
drop table t2;
drop table t3_null;
--
-- End of Test Parallel Hash Left Anti Semi (Not-In) Join.
--
--
-- Test alter ao/aocs table parallel_workers options
--
begin;
Expand Down
33 changes: 33 additions & 0 deletions src/test/regress/sql/gp_parallel.sql
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -465,6 +465,39 @@ set local enable_parallel = off;
explain(costs off, locus) select * from t1 order by c2 asc limit 3 offset 5;
select * from t1 order by c2 asc limit 3 offset 5;
abort;

--
-- Test Parallel Hash Left Anti Semi (Not-In) Join(parallel-oblivious).
--
create table t1(c1 int, c2 int) using ao_row distributed by (c1);
create table t2(c1 int, c2 int) using ao_row distributed by (c1);
create table t3_null(c1 int, c2 int) using ao_row distributed by (c1);
set enable_parallel = on;
set gp_appendonly_insert_files = 2;
set gp_appendonly_insert_files_tuples_range = 100;
set max_parallel_workers_per_gather = 2;
insert into t1 select i, i from generate_series(1, 5000000) i;
insert into t2 select i+1, i from generate_series(1, 1200) i;
insert into t3_null select i+1, i from generate_series(1, 1200) i;
insert into t3_null values(NULL, NULL);
analyze t1;
analyze t2;
analyze t3_null;
explain(costs off) select sum(t1.c1) from t1 where c1 not in (select c1 from t2);
select sum(t1.c1) from t1 where c1 not in (select c1 from t2);
explain(costs off) select * from t1 where c1 not in (select c1 from t3_null);
select * from t1 where c1 not in (select c1 from t3_null);
-- non-parallel results.
set enable_parallel = off;
select sum(t1.c1) from t1 where c1 not in (select c1 from t2);
select * from t1 where c1 not in (select c1 from t3_null);
drop table t1;
drop table t2;
drop table t3_null;
--
-- End of Test Parallel Hash Left Anti Semi (Not-In) Join.
--

--
-- Test alter ao/aocs table parallel_workers options
--
Expand Down