mirror of
https://github.com/timescale/timescaledb.git
synced 2025-05-17 19:13:16 +08:00
00a096f2f6
When testing on Windows these tests were returning differences because of floating point (sql_query_results_*) differences and because the cost was trivially higher (insert_single). Here we remove the costs since it was not relevant to the regressions we want to detect. Further we truncate the floating point averages to 8 decimal places; enough to see significant differences.
198 lines
9.0 KiB
PL/PgSQL
198 lines
9.0 KiB
PL/PgSQL
CREATE TABLE PUBLIC.hyper_1 (
|
|
time TIMESTAMP NOT NULL,
|
|
series_0 DOUBLE PRECISION NULL,
|
|
series_1 DOUBLE PRECISION NULL,
|
|
series_2 DOUBLE PRECISION NULL
|
|
);
|
|
|
|
CREATE INDEX "time_plain" ON PUBLIC.hyper_1 (time DESC, series_0);
|
|
SELECT * FROM create_hypertable('"public"."hyper_1"'::regclass, 'time'::name, number_partitions => 1, create_default_indexes=>false);
|
|
INSERT INTO hyper_1 SELECT to_timestamp(ser), ser, ser+10000, sqrt(ser::numeric) FROM generate_series(0,10000) ser;
|
|
INSERT INTO hyper_1 SELECT to_timestamp(ser), ser, ser+10000, sqrt(ser::numeric) FROM generate_series(10001,20000) ser;
|
|
|
|
CREATE TABLE PUBLIC.hyper_1_tz (
|
|
time TIMESTAMPTZ NOT NULL,
|
|
series_0 DOUBLE PRECISION NULL,
|
|
series_1 DOUBLE PRECISION NULL,
|
|
series_2 DOUBLE PRECISION NULL
|
|
);
|
|
|
|
CREATE INDEX "time_plain_tz" ON PUBLIC.hyper_1_tz (time DESC, series_0);
|
|
SELECT * FROM create_hypertable('"public"."hyper_1_tz"'::regclass, 'time'::name, number_partitions => 1, create_default_indexes=>false);
|
|
INSERT INTO hyper_1_tz SELECT to_timestamp(ser), ser, ser+10000, sqrt(ser::numeric) FROM generate_series(0,10000) ser;
|
|
INSERT INTO hyper_1_tz SELECT to_timestamp(ser), ser, ser+10000, sqrt(ser::numeric) FROM generate_series(10001,20000) ser;
|
|
|
|
CREATE TABLE PUBLIC.hyper_1_int (
|
|
time int NOT NULL,
|
|
series_0 DOUBLE PRECISION NULL,
|
|
series_1 DOUBLE PRECISION NULL,
|
|
series_2 DOUBLE PRECISION NULL
|
|
);
|
|
|
|
CREATE INDEX "time_plain_int" ON PUBLIC.hyper_1_int (time DESC, series_0);
|
|
SELECT * FROM create_hypertable('"public"."hyper_1_int"'::regclass, 'time'::name, number_partitions => 1, chunk_time_interval=>10000, create_default_indexes=>FALSE);
|
|
INSERT INTO hyper_1_int SELECT ser, ser, ser+10000, sqrt(ser::numeric) FROM generate_series(0,10000) ser;
|
|
INSERT INTO hyper_1_int SELECT ser, ser, ser+10000, sqrt(ser::numeric) FROM generate_series(10001,20000) ser;
|
|
|
|
CREATE TABLE PUBLIC.hyper_1_date (
|
|
time date NOT NULL,
|
|
series_0 DOUBLE PRECISION NULL,
|
|
series_1 DOUBLE PRECISION NULL,
|
|
series_2 DOUBLE PRECISION NULL
|
|
);
|
|
|
|
CREATE INDEX "time_plain_date" ON PUBLIC.hyper_1_date (time DESC, series_0);
|
|
SELECT * FROM create_hypertable('"public"."hyper_1_date"'::regclass, 'time'::name, number_partitions => 1, chunk_time_interval=>86400000000, create_default_indexes=>FALSE);
|
|
INSERT INTO hyper_1_date SELECT to_timestamp(ser)::date, ser, ser+10000, sqrt(ser::numeric) FROM generate_series(0,10000) ser;
|
|
INSERT INTO hyper_1_date SELECT to_timestamp(ser)::date, ser, ser+10000, sqrt(ser::numeric) FROM generate_series(10001,20000) ser;
|
|
--below needed to create enough unique dates to trigger an index scan
|
|
INSERT INTO hyper_1_date SELECT to_timestamp(ser*100)::date, ser, ser+10000, sqrt(ser::numeric) FROM generate_series(10001,20000) ser;
|
|
|
|
CREATE TABLE PUBLIC.plain_table (
|
|
time TIMESTAMPTZ NOT NULL,
|
|
series_0 DOUBLE PRECISION NULL,
|
|
series_1 DOUBLE PRECISION NULL,
|
|
series_2 DOUBLE PRECISION NULL
|
|
);
|
|
|
|
CREATE INDEX "time_plain_plain_table" ON PUBLIC.plain_table (time DESC, series_0);
|
|
INSERT INTO plain_table SELECT to_timestamp(ser), ser, ser+10000, sqrt(ser::numeric) FROM generate_series(0,10000) ser;
|
|
INSERT INTO plain_table SELECT to_timestamp(ser), ser, ser+10000, sqrt(ser::numeric) FROM generate_series(10001,20000) ser;
|
|
|
|
SET client_min_messages = 'error';
|
|
--avoid warning polluting output
|
|
ANALYZE;
|
|
RESET client_min_messages;
|
|
|
|
|
|
|
|
--non-aggregates use MergeAppend in both optimized and non-optimized
|
|
EXPLAIN (costs off) SELECT * FROM hyper_1 ORDER BY "time" DESC limit 2;
|
|
SELECT * FROM hyper_1 ORDER BY "time" DESC limit 2;
|
|
|
|
--aggregates use MergeAppend only in optimized
|
|
EXPLAIN (costs off) SELECT date_trunc('minute', time) t, avg(series_0), min(series_1), avg(series_2) FROM hyper_1 GROUP BY t ORDER BY t DESC limit 2;
|
|
|
|
EXPLAIN (costs off) SELECT date_trunc('minute', time) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM hyper_1_date GROUP BY t ORDER BY t DESC limit 2;
|
|
|
|
|
|
--the minute and second results should be diff
|
|
SELECT date_trunc('minute', time) t, avg(series_0), min(series_1), avg(series_2) FROM hyper_1 GROUP BY t ORDER BY t DESC limit 2;
|
|
SELECT date_trunc('second', time) t, avg(series_0), min(series_1), avg(series_2) FROM hyper_1 GROUP BY t ORDER BY t DESC limit 2;
|
|
|
|
--test that when index on time used by constraint, still works correctly
|
|
EXPLAIN (costs off)
|
|
SELECT date_trunc('minute', time) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM hyper_1
|
|
WHERE time < to_timestamp(900)
|
|
GROUP BY t
|
|
ORDER BY t DESC
|
|
LIMIT 2;
|
|
|
|
SELECT date_trunc('minute', time) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM hyper_1
|
|
WHERE time < to_timestamp(900)
|
|
GROUP BY t
|
|
ORDER BY t DESC
|
|
LIMIT 2;
|
|
|
|
--test that still works with an expression index on data_trunc.
|
|
DROP INDEX "time_plain";
|
|
CREATE INDEX "time_trunc" ON PUBLIC.hyper_1 (date_trunc('minute', time));
|
|
SET client_min_messages = 'error';
|
|
--avoid warning polluting output
|
|
ANALYZE;
|
|
RESET client_min_messages;
|
|
|
|
EXPLAIN (costs off) SELECT date_trunc('minute', time) t, avg(series_0), min(series_1), avg(series_2) FROM hyper_1 GROUP BY t ORDER BY t DESC limit 2;
|
|
SELECT date_trunc('minute', time) t, avg(series_0), min(series_1), avg(series_2) FROM hyper_1 GROUP BY t ORDER BY t DESC limit 2;
|
|
|
|
--test that works with both indexes
|
|
CREATE INDEX "time_plain" ON PUBLIC.hyper_1 (time DESC, series_0);
|
|
SET client_min_messages = 'error';
|
|
--avoid warning polluting output
|
|
ANALYZE;
|
|
RESET client_min_messages;
|
|
EXPLAIN (costs off) SELECT date_trunc('minute', time) t, avg(series_0), min(series_1), avg(series_2) FROM hyper_1 GROUP BY t ORDER BY t DESC limit 2;
|
|
SELECT date_trunc('minute', time) t, avg(series_0), min(series_1), avg(series_2) FROM hyper_1 GROUP BY t ORDER BY t DESC limit 2;
|
|
|
|
EXPLAIN (costs off) SELECT time_bucket('1 minute', time) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM hyper_1 GROUP BY t ORDER BY t DESC limit 2;
|
|
SELECT time_bucket('1 minute', time) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM hyper_1 GROUP BY t ORDER BY t DESC limit 2;
|
|
|
|
EXPLAIN (costs off) SELECT time_bucket('1 minute', time, INTERVAL '30 seconds') t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM hyper_1 GROUP BY t ORDER BY t DESC limit 2;
|
|
SELECT time_bucket('1 minute', time, INTERVAL '30 seconds') t, trunc(avg(series_0)::numeric, 8) as avg_trunc1, min(series_1), trunc(avg(series_2)::numeric, 8) as avg_trunc2
|
|
FROM hyper_1 GROUP BY t ORDER BY t DESC limit 2;
|
|
|
|
EXPLAIN (costs off) SELECT time_bucket('1 minute', time - INTERVAL '30 seconds') t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM hyper_1 GROUP BY t ORDER BY t DESC limit 2;
|
|
SELECT time_bucket('1 minute', time - INTERVAL '30 seconds') t, trunc(avg(series_0)::numeric, 8) as avg_trunc1, min(series_1), trunc(avg(series_2)::numeric, 8) as avg_trunc2
|
|
FROM hyper_1 GROUP BY t ORDER BY t DESC limit 2;
|
|
|
|
|
|
EXPLAIN (costs off) SELECT time_bucket('1 minute', time - INTERVAL '30 seconds') + INTERVAL '30 seconds' t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM hyper_1 GROUP BY t ORDER BY t DESC limit 2;
|
|
SELECT time_bucket('1 minute', time - INTERVAL '30 seconds') + INTERVAL '30 seconds' t, trunc(avg(series_0)::numeric, 8) as avg_trunc1, min(series_1), trunc(avg(series_2)::numeric, 8) as avg_trunc2
|
|
FROM hyper_1 GROUP BY t ORDER BY t DESC limit 2;
|
|
|
|
|
|
EXPLAIN (costs off) SELECT time_bucket('1 minute', time) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM hyper_1_tz GROUP BY t ORDER BY t DESC limit 2;
|
|
SELECT time_bucket('1 minute', time) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM hyper_1_tz GROUP BY t ORDER BY t DESC limit 2;
|
|
|
|
EXPLAIN (costs off) SELECT time_bucket('1 minute', time::timestamp) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM hyper_1_tz GROUP BY t ORDER BY t DESC limit 2;
|
|
SELECT time_bucket('1 minute', time::timestamp) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM hyper_1_tz GROUP BY t ORDER BY t DESC limit 2;
|
|
|
|
EXPLAIN (costs off) SELECT time_bucket(10, time) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM hyper_1_int GROUP BY t ORDER BY t DESC limit 2;
|
|
SELECT time_bucket(10, time) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM hyper_1_int GROUP BY t ORDER BY t DESC limit 2;
|
|
|
|
EXPLAIN (costs off) SELECT time_bucket(10, time, 2) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM hyper_1_int GROUP BY t ORDER BY t DESC limit 2;
|
|
SELECT time_bucket(10, time) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM hyper_1_int GROUP BY t ORDER BY t DESC limit 2;
|
|
|
|
|
|
--plain tables shouldnt be optimized by default
|
|
EXPLAIN (costs off)
|
|
SELECT date_trunc('minute', time) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM plain_table
|
|
WHERE time < to_timestamp(900)
|
|
GROUP BY t
|
|
ORDER BY t DESC
|
|
LIMIT 2;
|
|
|
|
SELECT date_trunc('minute', time) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM plain_table
|
|
WHERE time < to_timestamp(900)
|
|
GROUP BY t
|
|
ORDER BY t DESC
|
|
LIMIT 2;
|
|
|
|
|
|
--can turn on plain table optimizations
|
|
BEGIN;
|
|
SET LOCAL timescaledb.optimize_non_hypertables = 'on';
|
|
EXPLAIN (costs off)
|
|
SELECT date_trunc('minute', time) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM plain_table
|
|
WHERE time < to_timestamp(900)
|
|
GROUP BY t
|
|
ORDER BY t DESC
|
|
LIMIT 2;
|
|
|
|
SELECT date_trunc('minute', time) t, avg(series_0), min(series_1), avg(series_2)
|
|
FROM plain_table
|
|
WHERE time < to_timestamp(900)
|
|
GROUP BY t
|
|
ORDER BY t DESC
|
|
LIMIT 2;
|
|
ROLLBACK;
|