Skip to content
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

Jump to bottom

feat: store certain meta statistics about db #471

Draft
wants to merge 2 commits into
base: main
Choose a base branch
from
Draft

feat: store certain meta statistics about db #471

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
7a0e6b8
chore: ignore default docker env file
tazlin Nov 5, 2024
d5a7ad6
feat: store certain meta statistics about db
tazlin Nov 5, 2024
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
1 change: 1 addition & 0 deletions .gitignore
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -147,3 +147,4 @@ apikey.txt
.vscode/

horde_sdk**
.env_docker
1 change: 1 addition & 0 deletions horde/classes/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -47,6 +47,7 @@
"cron/", # Must be first
"stored_procedures/",
"stored_procedures/cron_jobs/",
"meta_statistics/",
]

all_dirs_to_run = [sql_statement_dir / dir for dir in all_dirs_to_run]
Expand Down
271 changes: 271 additions & 0 deletions sql_statements/meta_statistics/meta_statistics.sql
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,271 @@
/*
SPDX-FileCopyrightText: 2023 Tazlin

SPDX-License-Identifier: AGPL-3.0-or-later
*/

CREATE SCHEMA IF NOT EXISTS statistics;

/* Create 'readonly' role and grant SELECT on tables in the 'statistics' schema
CREATE ROLE readonly;
GRANT CONNECT ON DATABASE postgres TO readonly;
GRANT USAGE ON SCHEMA statistics TO readonly;
GRANT SELECT ON ALL TABLES IN SCHEMA statistics TO readonly;
ALTER DEFAULT PRIVILEGES IN SCHEMA statistics GRANT SELECT ON TABLES to readonly;
*/

/* Waiting connections statistics

Original query:
SELECT (NOW() at time zone 'utc') AS time,
COUNT(*) AS waiting_connections
FROM pg_stat_activity
WHERE wait_event_type = 'Lock';

*/
CREATE TABLE IF NOT EXISTS statistics.waiting_connections (
time TIMESTAMP,
waiting_connections INTEGER
);

CREATE INDEX IF NOT EXISTS idx_waiting_connections_time
ON statistics.waiting_connections(time);

CREATE OR REPLACE PROCEDURE log_waiting_connections()
LANGUAGE plpgsql
AS $$
DECLARE
waiting_count INT;
BEGIN
-- Get the count of waiting connections
SELECT COUNT(*) INTO waiting_count
FROM pg_stat_activity
WHERE wait_event_type = 'Lock';

-- Insert only if waiting_count is greater than 0
IF waiting_count > 0 THEN
INSERT INTO statistics.waiting_connections (time, waiting_connections)
VALUES ((NOW() at time zone 'utc'), waiting_count);
END IF;
END;
$$;

CALL schedule_cron_job('0-59 * * * *', 'log_waiting_connections');


/* Bucketed query statistics
Original query:
SELECT (NOW() at time zone 'utc') AS time,
state,
query_duration_bucket,
MAX(long_running_queries) AS long_running_queries
FROM (
SELECT CASE
WHEN (clock_timestamp() - query_start) < interval '2 seconds'
THEN '1-2 seconds'
WHEN (clock_timestamp() - query_start) < interval '5 seconds'
THEN '2-5 seconds'
WHEN (clock_timestamp() - query_start) < interval '10 seconds'
THEN '5-10 seconds'
WHEN (clock_timestamp() - query_start) < interval '30 seconds'
THEN '10-30 seconds'
WHEN (clock_timestamp() - query_start) < interval '60 seconds'
THEN '30-60 seconds'
WHEN (clock_timestamp() - query_start) < interval '300 seconds'
THEN '60-300 seconds'
ELSE '300+ seconds'
END AS query_duration_bucket,
COUNT(*) AS long_running_queries,
state
FROM pg_stat_activity
WHERE (clock_timestamp() - query_start) > interval '1 second'
GROUP BY query_duration_bucket, state
) AS query_duration_bucket
WHERE long_running_queries > 0
GROUP BY query_duration_bucket, state
ORDER BY state,
CASE query_duration_bucket
WHEN '1-2 seconds' THEN 1
WHEN '2-5 seconds' THEN 2
WHEN '5-10 seconds' THEN 3
WHEN '10-30 seconds' THEN 4
WHEN '30-60 seconds' THEN 5
WHEN '60-300 seconds' THEN 6
WHEN '300+ seconds' THEN 7
END;

*/

CREATE TABLE IF NOT EXISTS statistics.bucketed_query_statistics (
time TIMESTAMP,
state TEXT,
query_duration_bucket TEXT,
long_running_queries INTEGER
);

CREATE INDEX IF NOT EXISTS idx_bucketed_query_statistics_time
ON statistics.bucketed_query_statistics(time);

CREATE INDEX IF NOT EXISTS idx_bucketed_query_statistics_time_state
ON statistics.bucketed_query_statistics(time, state);

CREATE INDEX IF NOT EXISTS idx_bucketed_query_statistics_duration_bucket
ON statistics.bucketed_query_statistics(query_duration_bucket);

CREATE INDEX IF NOT EXISTS idx_bucketed_query_statistics_duration_bucket_state
ON statistics.bucketed_query_statistics(query_duration_bucket, state);

CREATE OR REPLACE PROCEDURE log_bucketed_query_statistics()
LANGUAGE plpgsql
AS $$
BEGIN
INSERT INTO statistics.bucketed_query_statistics (time, state, query_duration_bucket, long_running_queries)
SELECT (NOW() at time zone 'utc') AS time,
state,
query_duration_bucket,
MAX(long_running_queries) AS long_running_queries
FROM (
SELECT CASE
WHEN (clock_timestamp() - query_start) < interval '2 seconds'
THEN '1-2 seconds'
WHEN (clock_timestamp() - query_start) < interval '5 seconds'
THEN '2-5 seconds'
WHEN (clock_timestamp() - query_start) < interval '10 seconds'
THEN '5-10 seconds'
WHEN (clock_timestamp() - query_start) < interval '30 seconds'
THEN '10-30 seconds'
WHEN (clock_timestamp() - query_start) < interval '60 seconds'
THEN '30-60 seconds'
WHEN (clock_timestamp() - query_start) < interval '300 seconds'
THEN '60-300 seconds'
ELSE '300+ seconds'
END AS query_duration_bucket,
COUNT(*) AS long_running_queries,
state
FROM pg_stat_activity
WHERE (clock_timestamp() - query_start) > interval '1 second'
GROUP BY query_duration_bucket, state
) AS query_duration_bucket
WHERE long_running_queries > 0
GROUP BY query_duration_bucket, state
ORDER BY state,
CASE query_duration_bucket
WHEN '1-2 seconds' THEN 1
WHEN '2-5 seconds' THEN 2
WHEN '5-10 seconds' THEN 3
WHEN '10-30 seconds' THEN 4
WHEN '30-60 seconds' THEN 5
WHEN '60-300 seconds' THEN 6
WHEN '300+ seconds' THEN 7
END;
END;
$$;

CALL schedule_cron_job('0-59 * * * *', 'log_bucketed_query_statistics');

/* Lock conflicts statistics
Original query:
SELECT (NOW() at time zone 'utc') AS time,
COUNT(*) AS lock_conflicts
FROM pg_stat_activity t1
JOIN pg_locks l1 ON t1.pid = l1.pid
JOIN pg_locks l2 ON l1.locktype = l2.locktype
AND l1.DATABASE IS NOT DISTINCT FROM l2.DATABASE
AND l1.relation IS NOT DISTINCT FROM l2.relation
WHERE t1.state = 'active'
AND l1.granted IS NOT true;

*/
CREATE TABLE IF NOT EXISTS statistics.lock_conflicts (
time TIMESTAMP,
lock_conflicts INTEGER
);

CREATE INDEX IF NOT EXISTS idx_lock_conflicts_time
ON statistics.lock_conflicts(time);

CREATE OR REPLACE PROCEDURE log_lock_conflicts()
LANGUAGE plpgsql
AS $$
BEGIN
INSERT INTO statistics.lock_conflicts (time, lock_conflicts)
SELECT (NOW() at time zone 'utc') AS time,
COUNT(*) AS lock_conflicts
FROM pg_stat_activity t1
JOIN pg_locks l1 ON t1.pid = l1.pid
JOIN pg_locks l2 ON l1.locktype = l2.locktype
AND l1.DATABASE IS NOT DISTINCT FROM l2.DATABASE
AND l1.relation IS NOT DISTINCT FROM l2.relation
WHERE t1.state = 'active'
AND l1.granted IS NOT true;
END;
$$;

CALL schedule_cron_job('0-59 * * * *', 'log_lock_conflicts');

/* Dead row statistics
Original query:
SELECT (NOW() at time zone 'utc') AS time,
SUM(COALESCE(n_dead_tup,0)) AS dead_rows
FROM pg_stat_user_tables;
*/
CREATE TABLE IF NOT EXISTS statistics.dead_rows (
time TIMESTAMP,
dead_rows INTEGER
);

CREATE INDEX IF NOT EXISTS idx_dead_rows_time
ON statistics.dead_rows(time);

CREATE OR REPLACE PROCEDURE log_dead_rows()
LANGUAGE plpgsql
AS $$
BEGIN
INSERT INTO statistics.dead_rows (time, dead_rows)
SELECT (NOW() at time zone 'utc') AS time,
SUM(COALESCE(n_dead_tup,0)) AS dead_rows
FROM pg_stat_user_tables;
END;
$$;

CALL schedule_cron_job('0-59 * * * *', 'log_dead_rows');

UPDATE cron.job SET nodename = '';


/* Any-state long-running-queries count statistics
Original query:
SELECT COUNT(*)
FROM pg_stat_activity
WHERE
state IS NOT NULL
AND query NOT LIKE '%ROLLBACK%'
AND query NOT LIKE '%COMMIT%'
AND (now() - query_start) > interval '5 seconds';
*/
CREATE TABLE IF NOT EXISTS statistics.any_state_long_running_queries (
time TIMESTAMP,
long_running_queries INTEGER
);

CREATE INDEX IF NOT EXISTS idx_any_state_long_running_queries_time
ON statistics.any_state_long_running_queries(time);

CREATE OR REPLACE PROCEDURE log_any_state_long_running_queries()
LANGUAGE plpgsql
AS $$
BEGIN
INSERT INTO statistics.any_state_long_running_queries (time, long_running_queries)
SELECT (NOW() at time zone 'utc') AS time,
COUNT(*)
FROM pg_stat_activity
WHERE
state IS NOT NULL
AND query NOT LIKE '%ROLLBACK%'
AND query NOT LIKE '%COMMIT%'
AND (now() - query_start) > interval '5 seconds';
END;

$$;

CALL schedule_cron_job('0-59 * * * *', 'log_any_state_long_running_queries');
Loading