[MDEV-41399] With Parallel execution Performance degradation seen in JOIN query - Jira

XML

Word

Printable

Huge performance drop seen in JOIN query:

Query:
select * from s1 join s2 on(1);

Serial execution: 0.19sec
Parallel execution : 2.167sec

For analyze output refer:
serial : analyze-json.txt
Parallel : analyze-json-PQ.txt

----------------------
Branch: bb-13.2-MDEV-40012-refactored-input3
Commit:fcf80b7794ee57993cf7001a288a53b47a7e74ed
Version:13.1.0

CREATE TABLE s1 (a INT PRIMARY KEY, b INT, c VARCHAR(50), d DECIMAL(10,2), KEY(b)) ENGINE=InnoDB;

INSERT INTO s1 SELECT seq, seq%100, CONCAT('v',seq%7), seq/3 FROM seq_1_to_1000;

ANALYZE TABLE s1;

CREATE TABLE s2 (a INT PRIMARY KEY, b INT, c VARCHAR(50), d DECIMAL(10,2), KEY(b)) ENGINE=InnoDB;

INSERT INTO s2 SELECT seq, seq%100, CONCAT('v',seq%7), seq/3 FROM seq_1_to_2000;

ANALYZE TABLE s2;

SET SESSION parallel_worker_threads = 0;

select * from s1 join s2 on(1);

SET SESSION parallel_worker_threads = 4;

select * from s1 join s2 on(1);

is caused by

MDEV-40012 Parallel Query: execute the join in the worker threads

Estimated:

Remaining:

Logged:

Error rendering 'com.xiplink.jira.git.jira_git_plugin:git-issue-webpanel'. Please contact your Jira administrators.