Movatterモバイル変換

[0]ホーム
URL:

Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

 Sign up
Appearance settings

Commitbda6e5b

Browse files
author
Maksim Milyutin
committed
Fix tests for parallel queries
1 parenteffb898 commitbda6e5b

File tree

1 file changed

+131
-24
lines changed

1 file changed

+131
-24
lines changed

‎tests/partitioning_test.py‎

Lines changed: 131 additions & 24 deletions
Original file line numberDiff line numberDiff line change
@@ -409,6 +409,8 @@ def test_foreign_table(self):
409409
deftest_parallel_nodes(self):
410410
"""Test parallel queries under partitions"""
411411

412+
importjson
413+
412414
# Init and start postgres instance with preload pg_pathman module
413415
node=get_new_node('test')
414416
node.init()
@@ -419,8 +421,8 @@ def test_parallel_nodes(self):
419421

420422
# Check version of postgres server
421423
# If version < 9.6 skip all tests for parallel queries
422-
version=node.psql("postgres","show server_version_num")
423-
ifint(version[1])<90600:
424+
version=int(node.psql("postgres","show server_version_num")[1])
425+
ifversion<90600:
424426
return
425427

426428
# Prepare test database
@@ -435,6 +437,26 @@ def test_parallel_nodes(self):
435437
node.psql('postgres','select create_hash_partitions(\'hash_partitioned\',\'i\', 10)')
436438
node.psql('postgres','vacuum analyze hash_partitioned')
437439

440+
node.psql('postgres',"""
441+
create or replace function query_plan(query text) returns jsonb as $$
442+
declare
443+
plan jsonb;
444+
begin
445+
execute 'explain (costs off, format json)' || query into plan;
446+
return plan;
447+
end;
448+
$$ language plpgsql;
449+
""")
450+
451+
# Helper function for json equality
452+
defordered(obj):
453+
ifisinstance(obj,dict):
454+
returnsorted((k,ordered(v))fork,vinobj.items())
455+
ifisinstance(obj,list):
456+
returnsorted(ordered(x)forxinobj)
457+
else:
458+
returnobj
459+
438460
# Test parallel select
439461
withnode.connect()ascon:
440462
con.execute('set max_parallel_workers_per_gather = 2')
@@ -443,35 +465,120 @@ def test_parallel_nodes(self):
443465
con.execute('set parallel_tuple_cost = 0')
444466

445467
# Check parallel aggregate plan
446-
plan=con.execute('explain (costs off) select count(*) from range_partitioned where i < 1500')
447-
expected= [('Finalize Aggregate',),
448-
(' -> Gather',),
449-
(' Workers Planned: 2',),
450-
(' -> Partial Aggregate',),
451-
(' -> Append',),
452-
(' -> Parallel Seq Scan on range_partitioned_1',),
453-
(' -> Parallel Seq Scan on range_partitioned_2',),
454-
(' Filter: (i < 1500)',)]
455-
self.assertEqual(plan,expected)
468+
test_query='select count(*) from range_partitioned where i < 1500'
469+
plan=con.execute('select query_plan(\'%s\')'%test_query)[0][0]
470+
expected=json.loads("""
471+
[
472+
{
473+
"Plan": {
474+
"Node Type": "Aggregate",
475+
"Strategy": "Plain",
476+
"Partial Mode": "Finalize",
477+
"Parallel Aware": false,
478+
"Plans": [
479+
{
480+
"Node Type": "Gather",
481+
"Parent Relationship": "Outer",
482+
"Parallel Aware": false,
483+
"Workers Planned": 2,
484+
"Single Copy": false,
485+
"Plans": [
486+
{
487+
"Node Type": "Aggregate",
488+
"Strategy": "Plain",
489+
"Partial Mode": "Partial",
490+
"Parent Relationship": "Outer",
491+
"Parallel Aware": false,
492+
"Plans": [
493+
{
494+
"Node Type": "Append",
495+
"Parent Relationship": "Outer",
496+
"Parallel Aware": false,
497+
"Plans": [
498+
{
499+
"Node Type": "Seq Scan",
500+
"Parent Relationship": "Member",
501+
"Parallel Aware": true,
502+
"Relation Name": "range_partitioned_2",
503+
"Alias": "range_partitioned_2",
504+
"Filter": "(i < 1500)"
505+
},
506+
{
507+
"Node Type": "Seq Scan",
508+
"Parent Relationship": "Member",
509+
"Parallel Aware": true,
510+
"Relation Name": "range_partitioned_1",
511+
"Alias": "range_partitioned_1"
512+
}
513+
]
514+
}
515+
]
516+
}
517+
]
518+
}
519+
]
520+
}
521+
}
522+
]
523+
""")
524+
self.assertEqual(ordered(plan),ordered(expected))
456525

457526
# Check count of returned tuples
458-
count=con.execute('select count(*) from range_partitioned where i < 1500')
459-
self.assertEqual(count[0][0],1499)
527+
count=con.execute('select count(*) from range_partitioned where i < 1500')[0][0]
528+
self.assertEqual(count,1499)
460529

461530
# Check simple parallel seq scan plan with limit
462-
plan=con.execute('explain (costs off) select * from range_partitioned where i < 1500 limit 5')
463-
expected= [('Limit',),
464-
(' -> Gather',),
465-
(' Workers Planned: 2',),
466-
(' -> Append',),
467-
(' -> Parallel Seq Scan on range_partitioned_1',),
468-
(' -> Parallel Seq Scan on range_partitioned_2',),
469-
(' Filter: (i < 1500)',)]
470-
self.assertEqual(plan,expected)
531+
test_query='select * from range_partitioned where i < 1500 limit 5'
532+
plan=con.execute('select query_plan(\'%s\')'%test_query)[0][0]
533+
expected=json.loads("""
534+
[
535+
{
536+
"Plan": {
537+
"Node Type": "Limit",
538+
"Parallel Aware": false,
539+
"Plans": [
540+
{
541+
"Node Type": "Gather",
542+
"Parent Relationship": "Outer",
543+
"Parallel Aware": false,
544+
"Workers Planned": 2,
545+
"Single Copy": false,
546+
"Plans": [
547+
{
548+
"Node Type": "Append",
549+
"Parent Relationship": "Outer",
550+
"Parallel Aware": false,
551+
"Plans": [
552+
{
553+
"Node Type": "Seq Scan",
554+
"Parent Relationship": "Member",
555+
"Parallel Aware": true,
556+
"Relation Name": "range_partitioned_2",
557+
"Alias": "range_partitioned_2",
558+
"Filter": "(i < 1500)"
559+
},
560+
{
561+
"Node Type": "Seq Scan",
562+
"Parent Relationship": "Member",
563+
"Parallel Aware": true,
564+
"Relation Name": "range_partitioned_1",
565+
"Alias": "range_partitioned_1"
566+
}
567+
]
568+
}
569+
]
570+
}
571+
]
572+
}
573+
}
574+
]
575+
""")
576+
self.assertEqual(ordered(plan),ordered(expected))
471577

472578
# Check tuples returned by query above
473579
res_tuples=con.execute('select * from range_partitioned where i < 1500 limit 5')
474-
expected= [(1,), (2,), (3,), (4,), (5,)]
580+
res_tuples=sorted(map(lambdax:x[0],res_tuples))
581+
expected= [1,2,3,4,5]
475582
self.assertEqual(res_tuples,expected)
476583
# import ipdb; ipdb.set_trace()
477584

0 commit comments

Comments
 (0)
[8]ページ先頭
©2009-2025 Movatter.jp