Files
greptimedb/tests/cases/distributed/explain/step_aggr_basic.result
Yingwen b719c020ba chore: cherry pick #6540, #6550, #6551, #6556, #6563, #6534 to v0.15 branch (#6577)
* feat: add metrics for request wait time and adjust stall metrics (#6540)

* feat: add metric greptime_mito_request_wait_time to observe wait time

Signed-off-by: evenyag <realevenyag@gmail.com>

* feat: add worker to wait time metric

Signed-off-by: evenyag <realevenyag@gmail.com>

* refactor: rename stall gauge to greptime_mito_write_stalling_count

Signed-off-by: evenyag <realevenyag@gmail.com>

* feat: change greptime_mito_write_stall_total to total stalled requests

Signed-off-by: evenyag <realevenyag@gmail.com>

* refactor: merge lazy static blocks

Signed-off-by: evenyag <realevenyag@gmail.com>

---------

Signed-off-by: evenyag <realevenyag@gmail.com>

* fix: estimate mem size for bulk ingester (#6550)

Signed-off-by: evenyag <realevenyag@gmail.com>

* fix: flow mirror cache (#6551)

* fix: invalid cache when flownode change address

Signed-off-by: discord9 <discord9@163.com>

* update comments

Signed-off-by: discord9 <discord9@163.com>

* fix

Signed-off-by: discord9 <discord9@163.com>

* refactor: add log&rename

Signed-off-by: discord9 <discord9@163.com>

* stuff

Signed-off-by: discord9 <discord9@163.com>

---------

Signed-off-by: discord9 <discord9@163.com>
Signed-off-by: evenyag <realevenyag@gmail.com>

* feat: impl timestamp function for promql (#6556)

* feat: impl timestamp function for promql

Signed-off-by: Dennis Zhuang <killme2008@gmail.com>

* chore: style and typo

Signed-off-by: Dennis Zhuang <killme2008@gmail.com>

* fix: test

Signed-off-by: Dennis Zhuang <killme2008@gmail.com>

* docs: update comments

Signed-off-by: Dennis Zhuang <killme2008@gmail.com>

* chore: comment

Signed-off-by: Dennis Zhuang <killme2008@gmail.com>

---------

Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
Signed-off-by: evenyag <realevenyag@gmail.com>

* feat: MergeScan print input (#6563)

* feat: MergeScan print input

Signed-off-by: discord9 <discord9@163.com>

* test: fix ut

Signed-off-by: discord9 <discord9@163.com>

---------

Signed-off-by: discord9 <discord9@163.com>
Signed-off-by: evenyag <realevenyag@gmail.com>

* fix: aggr group by all partition cols use partial commutative (#6534)

* fix: aggr group by all partition cols use partial commutative

Signed-off-by: discord9 <discord9@163.com>

* test: bugged case

Signed-off-by: discord9 <discord9@163.com>

* test: sqlness fix

Signed-off-by: discord9 <discord9@163.com>

* test: more redacted

Signed-off-by: discord9 <discord9@163.com>

* more cases

Signed-off-by: discord9 <discord9@163.com>

* even more test cases

Signed-off-by: discord9 <discord9@163.com>

* join testcase

Signed-off-by: discord9 <discord9@163.com>

* fix: column requirement added in correct location

Signed-off-by: discord9 <discord9@163.com>

* fix test

Signed-off-by: discord9 <discord9@163.com>

* chore: clippy

Signed-off-by: discord9 <discord9@163.com>

* track col reqs per stack

Signed-off-by: discord9 <discord9@163.com>

* fix: continue

Signed-off-by: discord9 <discord9@163.com>

* chore: clippy

Signed-off-by: discord9 <discord9@163.com>

* refactor: test mod

Signed-off-by: discord9 <discord9@163.com>

* test utils

Signed-off-by: discord9 <discord9@163.com>

* test: better test

Signed-off-by: discord9 <discord9@163.com>

* more testcases

Signed-off-by: discord9 <discord9@163.com>

* test limit push down

Signed-off-by: discord9 <discord9@163.com>

* more testcases

Signed-off-by: discord9 <discord9@163.com>

* more testcase

Signed-off-by: discord9 <discord9@163.com>

* more test

Signed-off-by: discord9 <discord9@163.com>

* chore: update sqlness

Signed-off-by: discord9 <discord9@163.com>

* chore: update commnets

Signed-off-by: discord9 <discord9@163.com>

* fix: check col reqs from bottom to upper

Signed-off-by: discord9 <discord9@163.com>

* chore: more comment

Signed-off-by: discord9 <discord9@163.com>

* docs: more todo

Signed-off-by: discord9 <discord9@163.com>

* chore: comments

Signed-off-by: discord9 <discord9@163.com>

* test: a new failing test that should be fixed

Signed-off-by: discord9 <discord9@163.com>

* fix: part col alias tracking

Signed-off-by: discord9 <discord9@163.com>

* chore: unused

Signed-off-by: discord9 <discord9@163.com>

* chore: clippy

Signed-off-by: discord9 <discord9@163.com>

* docs: comment

Signed-off-by: discord9 <discord9@163.com>

* mroe testcase

Signed-off-by: discord9 <discord9@163.com>

* more testcase for step/part aggr combine

Signed-off-by: discord9 <discord9@163.com>

* FIXME: a new bug

Signed-off-by: discord9 <discord9@163.com>

* literally unfixable

Signed-off-by: discord9 <discord9@163.com>

* chore: remove some debug print

Signed-off-by: discord9 <discord9@163.com>

---------

Signed-off-by: discord9 <discord9@163.com>
Signed-off-by: evenyag <realevenyag@gmail.com>

---------

Signed-off-by: evenyag <realevenyag@gmail.com>
Signed-off-by: discord9 <discord9@163.com>
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
Co-authored-by: fys <40801205+fengys1996@users.noreply.github.com>
Co-authored-by: discord9 <55937128+discord9@users.noreply.github.com>
Co-authored-by: dennis zhuang <killme2008@gmail.com>
2025-07-23 22:29:14 +08:00

575 lines
23 KiB
Plaintext

CREATE TABLE integers(
host STRING PRIMARY KEY,
i BIGINT,
ts TIMESTAMP TIME INDEX
) PARTITION ON COLUMNS (host) (
host < '550-A',
host >= '550-A'
AND host < '550-W',
host >= '550-W'
);
Affected Rows: 0
INSERT INTO
integers (host, i, ts)
VALUES
('550-A', 1, '2023-01-01 00:00:00'),
('550-B', 5, '2023-01-01 00:00:00'),
('550-A', 2, '2023-01-01 01:00:00'),
('550-W', 3, '2023-01-01 02:00:00'),
('550-W', 4, '2023-01-01 03:00:00');
Affected Rows: 5
-- count
SELECT
count(i)
FROM
integers;
+-------------------+
| count(integers.i) |
+-------------------+
| 5 |
+-------------------+
-- SQLNESS REPLACE (-+) -
-- SQLNESS REPLACE (\s\s+) _
-- SQLNESS REPLACE (RoundRobinBatch.*) REDACTED
-- SQLNESS REPLACE (Hash.*) REDACTED
-- SQLNESS REPLACE (peers.*) REDACTED
EXPLAIN
SELECT
count(i)
FROM
integers;
+-+-+
| plan_type_| plan_|
+-+-+
| logical_plan_| Projection: sum(count(integers.i)) AS count(integers.i)_|
|_|_Aggregate: groupBy=[[]], aggr=[[sum(count(integers.i))]]_|
|_|_MergeScan [is_placeholder=false, remote_input=[_|
|_| Aggregate: groupBy=[[]], aggr=[[count(integers.i)]]_|
|_|_TableScan: integers_|
|_| ]]_|
| physical_plan | ProjectionExec: expr=[sum(count(integers.i))@0 as count(integers.i)]_|
|_|_AggregateExec: mode=Final, gby=[], aggr=[sum(count(integers.i))]_|
|_|_CoalescePartitionsExec_|
|_|_AggregateExec: mode=Partial, gby=[], aggr=[sum(count(integers.i))]_|
|_|_MergeScanExec: REDACTED
|_|_|
+-+-+
-- SQLNESS REPLACE (metrics.*) REDACTED
-- SQLNESS REPLACE (RoundRobinBatch.*) REDACTED
-- SQLNESS REPLACE (-+) -
-- SQLNESS REPLACE (\s\s+) _
-- SQLNESS REPLACE (peers.*) REDACTED
-- SQLNESS REPLACE region=\d+\(\d+,\s+\d+\) region=REDACTED
-- might write to different partitions
-- SQLNESS REPLACE "partition_count":\{(.*?)\} "partition_count":REDACTED
-- SQLNESS REPLACE (Hash.*) REDACTED
EXPLAIN ANALYZE
SELECT
count(i)
FROM
integers;
+-+-+-+
| stage | node | plan_|
+-+-+-+
| 0_| 0_|_ProjectionExec: expr=[sum(count(integers.i))@0 as count(integers.i)] REDACTED
|_|_|_AggregateExec: mode=Final, gby=[], aggr=[sum(count(integers.i))] REDACTED
|_|_|_CoalescePartitionsExec REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[], aggr=[sum(count(integers.i))] REDACTED
|_|_|_MergeScanExec: REDACTED
|_|_|_|
| 1_| 0_|_AggregateExec: mode=Final, gby=[], aggr=[count(integers.i)] REDACTED
|_|_|_CoalescePartitionsExec REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[], aggr=[count(integers.i)] REDACTED
|_|_|_SeqScan: region=REDACTED, "partition_count":REDACTED REDACTED
|_|_|_|
| 1_| 1_|_AggregateExec: mode=Final, gby=[], aggr=[count(integers.i)] REDACTED
|_|_|_CoalescePartitionsExec REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[], aggr=[count(integers.i)] REDACTED
|_|_|_SeqScan: region=REDACTED, "partition_count":REDACTED REDACTED
|_|_|_|
| 1_| 2_|_AggregateExec: mode=Final, gby=[], aggr=[count(integers.i)] REDACTED
|_|_|_CoalescePartitionsExec REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[], aggr=[count(integers.i)] REDACTED
|_|_|_SeqScan: region=REDACTED, "partition_count":REDACTED REDACTED
|_|_|_|
|_|_| Total rows: 1_|
+-+-+-+
SELECT
ts,
count(i)
FROM
integers
GROUP BY
ts
ORDER BY
ts,
count(i);
+---------------------+-------------------+
| ts | count(integers.i) |
+---------------------+-------------------+
| 2023-01-01T00:00:00 | 2 |
| 2023-01-01T01:00:00 | 1 |
| 2023-01-01T02:00:00 | 1 |
| 2023-01-01T03:00:00 | 1 |
+---------------------+-------------------+
-- SQLNESS REPLACE (-+) -
-- SQLNESS REPLACE (\s\s+) _
-- SQLNESS REPLACE (RoundRobinBatch.*) REDACTED
-- SQLNESS REPLACE (Hash.*) REDACTED
-- SQLNESS REPLACE (peers.*) REDACTED
EXPLAIN
SELECT
ts,
count(i)
FROM
integers
GROUP BY
ts
ORDER BY
ts,
count(i);
+-+-+
| plan_type_| plan_|
+-+-+
| logical_plan_| Sort: integers.ts ASC NULLS LAST, count(integers.i) ASC NULLS LAST_|
|_|_Projection: integers.ts, sum(count(integers.i)) AS count(integers.i)_|
|_|_Aggregate: groupBy=[[integers.ts]], aggr=[[sum(count(integers.i))]]_|
|_|_MergeScan [is_placeholder=false, remote_input=[_|
|_| Aggregate: groupBy=[[integers.ts]], aggr=[[count(integers.i)]]_|
|_|_TableScan: integers_|
|_| ]]_|
| physical_plan | SortPreservingMergeExec: [ts@0 ASC NULLS LAST, count(integers.i)@1 ASC NULLS LAST]_|
|_|_SortExec: expr=[ts@0 ASC NULLS LAST, count(integers.i)@1 ASC NULLS LAST], preserve_partitioning=[true]_|
|_|_ProjectionExec: expr=[ts@0 as ts, sum(count(integers.i))@1 as count(integers.i)]_|
|_|_AggregateExec: mode=FinalPartitioned, gby=[ts@0 as ts], aggr=[sum(count(integers.i))]_|
|_|_CoalesceBatchesExec: target_batch_size=8192_|
|_|_RepartitionExec: partitioning=REDACTED
|_|_AggregateExec: mode=Partial, gby=[ts@0 as ts], aggr=[sum(count(integers.i))]_|
|_|_MergeScanExec: REDACTED
|_|_|
+-+-+
-- SQLNESS REPLACE (metrics.*) REDACTED
-- SQLNESS REPLACE (RoundRobinBatch.*) REDACTED
-- SQLNESS REPLACE (-+) -
-- SQLNESS REPLACE (\s\s+) _
-- SQLNESS REPLACE (peers.*) REDACTED
-- SQLNESS REPLACE region=\d+\(\d+,\s+\d+\) region=REDACTED
-- might write to different partitions
-- SQLNESS REPLACE "partition_count":\{(.*?)\} "partition_count":REDACTED
-- SQLNESS REPLACE (Hash.*) REDACTED
EXPLAIN ANALYZE
SELECT
ts,
count(i)
FROM
integers
GROUP BY
ts
ORDER BY
ts,
count(i);
+-+-+-+
| stage | node | plan_|
+-+-+-+
| 0_| 0_|_SortPreservingMergeExec: [ts@0 ASC NULLS LAST, count(integers.i)@1 ASC NULLS LAST] REDACTED
|_|_|_SortExec: expr=[ts@0 ASC NULLS LAST, count(integers.i)@1 ASC NULLS LAST], preserve_partitioning=[true] REDACTED
|_|_|_ProjectionExec: expr=[ts@0 as ts, sum(count(integers.i))@1 as count(integers.i)] REDACTED
|_|_|_AggregateExec: mode=FinalPartitioned, gby=[ts@0 as ts], aggr=[sum(count(integers.i))] REDACTED
|_|_|_CoalesceBatchesExec: target_batch_size=8192 REDACTED
|_|_|_RepartitionExec: partitioning=REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[ts@0 as ts], aggr=[sum(count(integers.i))] REDACTED
|_|_|_MergeScanExec: REDACTED
|_|_|_|
| 1_| 0_|_AggregateExec: mode=FinalPartitioned, gby=[ts@0 as ts], aggr=[count(integers.i)] REDACTED
|_|_|_CoalesceBatchesExec: target_batch_size=8192 REDACTED
|_|_|_RepartitionExec: partitioning=REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[ts@1 as ts], aggr=[count(integers.i)] REDACTED
|_|_|_SeqScan: region=REDACTED, "partition_count":REDACTED REDACTED
|_|_|_|
| 1_| 1_|_AggregateExec: mode=FinalPartitioned, gby=[ts@0 as ts], aggr=[count(integers.i)] REDACTED
|_|_|_CoalesceBatchesExec: target_batch_size=8192 REDACTED
|_|_|_RepartitionExec: partitioning=REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[ts@1 as ts], aggr=[count(integers.i)] REDACTED
|_|_|_SeqScan: region=REDACTED, "partition_count":REDACTED REDACTED
|_|_|_|
| 1_| 2_|_AggregateExec: mode=FinalPartitioned, gby=[ts@0 as ts], aggr=[count(integers.i)] REDACTED
|_|_|_CoalesceBatchesExec: target_batch_size=8192 REDACTED
|_|_|_RepartitionExec: partitioning=REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[ts@1 as ts], aggr=[count(integers.i)] REDACTED
|_|_|_SeqScan: region=REDACTED, "partition_count":REDACTED REDACTED
|_|_|_|
|_|_| Total rows: 4_|
+-+-+-+
SELECT
date_bin('1 hour' :: INTERVAL, ts) as time_window,
count(i)
FROM
integers
GROUP BY
time_window
ORDER BY
time_window,
count(i);
+---------------------+-------------------+
| time_window | count(integers.i) |
+---------------------+-------------------+
| 2023-01-01T00:00:00 | 2 |
| 2023-01-01T01:00:00 | 1 |
| 2023-01-01T02:00:00 | 1 |
| 2023-01-01T03:00:00 | 1 |
+---------------------+-------------------+
-- SQLNESS REPLACE (-+) -
-- SQLNESS REPLACE (\s\s+) _
-- SQLNESS REPLACE (RoundRobinBatch.*) REDACTED
-- SQLNESS REPLACE (Hash.*) REDACTED
-- SQLNESS REPLACE (peers.*) REDACTED
EXPLAIN
SELECT
date_bin('1 hour' :: INTERVAL, ts) as time_window,
count(i)
FROM
integers
GROUP BY
time_window
ORDER BY
time_window,
count(i);
+-+-+
| plan_type_| plan_|
+-+-+
| logical_plan_| Sort: time_window ASC NULLS LAST, count(integers.i) ASC NULLS LAST_|
|_|_Projection: date_bin(Utf8("1 hour"),integers.ts) AS time_window, sum(count(integers.i)) AS count(integers.i)_|
|_|_Aggregate: groupBy=[[date_bin(Utf8("1 hour"),integers.ts)]], aggr=[[sum(count(integers.i))]]_|
|_|_MergeScan [is_placeholder=false, remote_input=[_|
|_| Aggregate: groupBy=[[date_bin(CAST(Utf8("1 hour") AS Interval(MonthDayNano)), integers.ts)]], aggr=[[count(integers.i)]]_|
|_|_TableScan: integers_|
|_| ]]_|
| physical_plan | SortPreservingMergeExec: [time_window@0 ASC NULLS LAST, count(integers.i)@1 ASC NULLS LAST]_|
|_|_SortExec: expr=[time_window@0 ASC NULLS LAST, count(integers.i)@1 ASC NULLS LAST], preserve_partitioning=[true]_|
|_|_ProjectionExec: expr=[date_bin(Utf8("1 hour"),integers.ts)@0 as time_window, sum(count(integers.i))@1 as count(integers.i)]_|
|_|_AggregateExec: mode=FinalPartitioned, gby=[date_bin(Utf8("1 hour"),integers.ts)@0 as date_bin(Utf8("1 hour"),integers.ts)], aggr=[sum(count(integers.i))] |
|_|_CoalesceBatchesExec: target_batch_size=8192_|
|_|_RepartitionExec: partitioning=REDACTED
|_|_AggregateExec: mode=Partial, gby=[date_bin(Utf8("1 hour"),integers.ts)@0 as date_bin(Utf8("1 hour"),integers.ts)], aggr=[sum(count(integers.i))]_|
|_|_MergeScanExec: REDACTED
|_|_|
+-+-+
-- SQLNESS REPLACE (metrics.*) REDACTED
-- SQLNESS REPLACE (RoundRobinBatch.*) REDACTED
-- SQLNESS REPLACE (-+) -
-- SQLNESS REPLACE (\s\s+) _
-- SQLNESS REPLACE (peers.*) REDACTED
-- SQLNESS REPLACE region=\d+\(\d+,\s+\d+\) region=REDACTED
-- might write to different partitions
-- SQLNESS REPLACE "partition_count":\{(.*?)\} "partition_count":REDACTED
-- SQLNESS REPLACE (Hash.*) REDACTED
EXPLAIN ANALYZE
SELECT
date_bin('1 hour' :: INTERVAL, ts) as time_window,
count(i)
FROM
integers
GROUP BY
time_window
ORDER BY
time_window,
count(i);
+-+-+-+
| stage | node | plan_|
+-+-+-+
| 0_| 0_|_SortPreservingMergeExec: [time_window@0 ASC NULLS LAST, count(integers.i)@1 ASC NULLS LAST] REDACTED
|_|_|_SortExec: expr=[time_window@0 ASC NULLS LAST, count(integers.i)@1 ASC NULLS LAST], preserve_partitioning=[true] REDACTED
|_|_|_ProjectionExec: expr=[date_bin(Utf8("1 hour"),integers.ts)@0 as time_window, sum(count(integers.i))@1 as count(integers.i)] REDACTED
|_|_|_AggregateExec: mode=FinalPartitioned, gby=[date_bin(Utf8("1 hour"),integers.ts)@0 as date_bin(Utf8("1 hour"),integers.ts)], aggr=[sum(count(integers.i))] REDACTED
|_|_|_CoalesceBatchesExec: target_batch_size=8192 REDACTED
|_|_|_RepartitionExec: partitioning=REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[date_bin(Utf8("1 hour"),integers.ts)@0 as date_bin(Utf8("1 hour"),integers.ts)], aggr=[sum(count(integers.i))] REDACTED
|_|_|_MergeScanExec: REDACTED
|_|_|_|
| 1_| 0_|_AggregateExec: mode=FinalPartitioned, gby=[date_bin(Utf8("1 hour"),integers.ts)@0 as date_bin(Utf8("1 hour"),integers.ts)], aggr=[count(integers.i)] REDACTED
|_|_|_CoalesceBatchesExec: target_batch_size=8192 REDACTED
|_|_|_RepartitionExec: partitioning=REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[date_bin(IntervalMonthDayNano { months: 0, days: 0, nanoseconds: 3600000000000 }, ts@1) as date_bin(Utf8("1 hour"),integers.ts)], aggr=[count(integers.i)] REDACTED
|_|_|_SeqScan: region=REDACTED, "partition_count":REDACTED REDACTED
|_|_|_|
| 1_| 1_|_AggregateExec: mode=FinalPartitioned, gby=[date_bin(Utf8("1 hour"),integers.ts)@0 as date_bin(Utf8("1 hour"),integers.ts)], aggr=[count(integers.i)] REDACTED
|_|_|_CoalesceBatchesExec: target_batch_size=8192 REDACTED
|_|_|_RepartitionExec: partitioning=REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[date_bin(IntervalMonthDayNano { months: 0, days: 0, nanoseconds: 3600000000000 }, ts@1) as date_bin(Utf8("1 hour"),integers.ts)], aggr=[count(integers.i)] REDACTED
|_|_|_SeqScan: region=REDACTED, "partition_count":REDACTED REDACTED
|_|_|_|
| 1_| 2_|_AggregateExec: mode=FinalPartitioned, gby=[date_bin(Utf8("1 hour"),integers.ts)@0 as date_bin(Utf8("1 hour"),integers.ts)], aggr=[count(integers.i)] REDACTED
|_|_|_CoalesceBatchesExec: target_batch_size=8192 REDACTED
|_|_|_RepartitionExec: partitioning=REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[date_bin(IntervalMonthDayNano { months: 0, days: 0, nanoseconds: 3600000000000 }, ts@1) as date_bin(Utf8("1 hour"),integers.ts)], aggr=[count(integers.i)] REDACTED
|_|_|_SeqScan: region=REDACTED, "partition_count":REDACTED REDACTED
|_|_|_|
|_|_| Total rows: 4_|
+-+-+-+
SELECT
ts::BIGINT + 1,
i / 2,
count(i)
FROM
integers
GROUP BY
ts::BIGINT + 1,
i / 2,
ORDER BY
ts::BIGINT + 1,
i / 2,
;
+------------------------+-----------------------+-------------------+
| integers.ts + Int64(1) | integers.i / Int64(2) | count(integers.i) |
+------------------------+-----------------------+-------------------+
| 1672531200001 | 0 | 1 |
| 1672531200001 | 2 | 1 |
| 1672534800001 | 1 | 1 |
| 1672538400001 | 1 | 1 |
| 1672542000001 | 2 | 1 |
+------------------------+-----------------------+-------------------+
-- SQLNESS REPLACE (-+) -
-- SQLNESS REPLACE (\s\s+) _
-- SQLNESS REPLACE (RoundRobinBatch.*) REDACTED
-- SQLNESS REPLACE (Hash.*) REDACTED
-- SQLNESS REPLACE (peers.*) REDACTED
EXPLAIN
SELECT
ts::BIGINT + 1,
i / 2,
count(i)
FROM
integers
GROUP BY
ts::BIGINT + 1,
i / 2,
ORDER BY
ts::BIGINT + 1,
i / 2,
;
+-+-+
| plan_type_| plan_|
+-+-+
| logical_plan_| Sort: integers.ts + Int64(1) ASC NULLS LAST, integers.i / Int64(2) ASC NULLS LAST_|
|_|_Projection: integers.ts + Int64(1), integers.i / Int64(2), sum(count(integers.i)) AS count(integers.i)_|
|_|_Aggregate: groupBy=[[integers.ts + Int64(1), integers.i / Int64(2)]], aggr=[[sum(count(integers.i))]]_|
|_|_MergeScan [is_placeholder=false, remote_input=[_|
|_| Aggregate: groupBy=[[CAST(integers.ts AS Int64) + Int64(1), integers.i / Int64(2)]], aggr=[[count(integers.i)]]_|
|_|_TableScan: integers_|
|_| ]]_|
| physical_plan | SortPreservingMergeExec: [integers.ts + Int64(1)@0 ASC NULLS LAST, integers.i / Int64(2)@1 ASC NULLS LAST]_|
|_|_SortExec: expr=[integers.ts + Int64(1)@0 ASC NULLS LAST, integers.i / Int64(2)@1 ASC NULLS LAST], preserve_partitioning=[true]_|
|_|_ProjectionExec: expr=[integers.ts + Int64(1)@0 as integers.ts + Int64(1), integers.i / Int64(2)@1 as integers.i / Int64(2), sum(count(integers.i))@2 as count(integers.i)]_|
|_|_AggregateExec: mode=FinalPartitioned, gby=[integers.ts + Int64(1)@0 as integers.ts + Int64(1), integers.i / Int64(2)@1 as integers.i / Int64(2)], aggr=[sum(count(integers.i))] |
|_|_CoalesceBatchesExec: target_batch_size=8192_|
|_|_RepartitionExec: partitioning=REDACTED
|_|_AggregateExec: mode=Partial, gby=[integers.ts + Int64(1)@0 as integers.ts + Int64(1), integers.i / Int64(2)@1 as integers.i / Int64(2)], aggr=[sum(count(integers.i))]_|
|_|_MergeScanExec: REDACTED
|_|_|
+-+-+
-- SQLNESS REPLACE (metrics.*) REDACTED
-- SQLNESS REPLACE (RoundRobinBatch.*) REDACTED
-- SQLNESS REPLACE (-+) -
-- SQLNESS REPLACE (\s\s+) _
-- SQLNESS REPLACE (peers.*) REDACTED
-- SQLNESS REPLACE region=\d+\(\d+,\s+\d+\) region=REDACTED
-- might write to different partitions
-- SQLNESS REPLACE "partition_count":\{(.*?)\} "partition_count":REDACTED
-- SQLNESS REPLACE (Hash.*) REDACTED
EXPLAIN ANALYZE
SELECT
ts::BIGINT + 1,
i / 2,
count(i)
FROM
integers
GROUP BY
ts::BIGINT + 1,
i / 2,
ORDER BY
ts::BIGINT + 1,
i / 2,
;
+-+-+-+
| stage | node | plan_|
+-+-+-+
| 0_| 0_|_SortPreservingMergeExec: [integers.ts + Int64(1)@0 ASC NULLS LAST, integers.i / Int64(2)@1 ASC NULLS LAST] REDACTED
|_|_|_SortExec: expr=[integers.ts + Int64(1)@0 ASC NULLS LAST, integers.i / Int64(2)@1 ASC NULLS LAST], preserve_partitioning=[true] REDACTED
|_|_|_ProjectionExec: expr=[integers.ts + Int64(1)@0 as integers.ts + Int64(1), integers.i / Int64(2)@1 as integers.i / Int64(2), sum(count(integers.i))@2 as count(integers.i)] REDACTED
|_|_|_AggregateExec: mode=FinalPartitioned, gby=[integers.ts + Int64(1)@0 as integers.ts + Int64(1), integers.i / Int64(2)@1 as integers.i / Int64(2)], aggr=[sum(count(integers.i))] REDACTED
|_|_|_CoalesceBatchesExec: target_batch_size=8192 REDACTED
|_|_|_RepartitionExec: partitioning=REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[integers.ts + Int64(1)@0 as integers.ts + Int64(1), integers.i / Int64(2)@1 as integers.i / Int64(2)], aggr=[sum(count(integers.i))] REDACTED
|_|_|_MergeScanExec: REDACTED
|_|_|_|
| 1_| 0_|_AggregateExec: mode=FinalPartitioned, gby=[integers.ts + Int64(1)@0 as integers.ts + Int64(1), integers.i / Int64(2)@1 as integers.i / Int64(2)], aggr=[count(integers.i)] REDACTED
|_|_|_CoalesceBatchesExec: target_batch_size=8192 REDACTED
|_|_|_RepartitionExec: partitioning=REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[CAST(ts@1 AS Int64) + 1 as integers.ts + Int64(1), i@0 / 2 as integers.i / Int64(2)], aggr=[count(integers.i)] REDACTED
|_|_|_SeqScan: region=REDACTED, "partition_count":REDACTED REDACTED
|_|_|_|
| 1_| 1_|_AggregateExec: mode=FinalPartitioned, gby=[integers.ts + Int64(1)@0 as integers.ts + Int64(1), integers.i / Int64(2)@1 as integers.i / Int64(2)], aggr=[count(integers.i)] REDACTED
|_|_|_CoalesceBatchesExec: target_batch_size=8192 REDACTED
|_|_|_RepartitionExec: partitioning=REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[CAST(ts@1 AS Int64) + 1 as integers.ts + Int64(1), i@0 / 2 as integers.i / Int64(2)], aggr=[count(integers.i)] REDACTED
|_|_|_SeqScan: region=REDACTED, "partition_count":REDACTED REDACTED
|_|_|_|
| 1_| 2_|_AggregateExec: mode=FinalPartitioned, gby=[integers.ts + Int64(1)@0 as integers.ts + Int64(1), integers.i / Int64(2)@1 as integers.i / Int64(2)], aggr=[count(integers.i)] REDACTED
|_|_|_CoalesceBatchesExec: target_batch_size=8192 REDACTED
|_|_|_RepartitionExec: partitioning=REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[CAST(ts@1 AS Int64) + 1 as integers.ts + Int64(1), i@0 / 2 as integers.i / Int64(2)], aggr=[count(integers.i)] REDACTED
|_|_|_SeqScan: region=REDACTED, "partition_count":REDACTED REDACTED
|_|_|_|
|_|_| Total rows: 5_|
+-+-+-+
-- test udd/hll_merege pushdown
CREATE TABLE sink_table (
time_window TIMESTAMP TIME INDEX,
host STRING PRIMARY KEY,
udd_state BINARY,
hll_state BINARY,
) PARTITION ON COLUMNS (host) (
host < '550-A',
host >= '550-A'
AND host < '550-W',
host >= '550-W'
);
Affected Rows: 0
INSERT INTO
sink_table
SELECT
date_bin('1 hour' :: INTERVAL, ts) as time_window,
host,
uddsketch_state(128, 0.01, i) as udd_state,
hll(i) as hll_state
FROM
integers
GROUP BY
time_window,
host;
Affected Rows: 5
SELECT
uddsketch_calc(0.5, uddsketch_merge(128, 0.01, udd_state)) as udd_result,
hll_count(hll_merge(hll_state)) as hll_result
FROM
sink_table;
+--------------------+------------+
| udd_result | hll_result |
+--------------------+------------+
| 2.9742334234767016 | 5 |
+--------------------+------------+
-- SQLNESS REPLACE (-+) -
-- SQLNESS REPLACE (\s\s+) _
-- SQLNESS REPLACE (RoundRobinBatch.*) REDACTED
-- SQLNESS REPLACE (Hash.*) REDACTED
-- SQLNESS REPLACE (peers.*) REDACTED
EXPLAIN
SELECT
uddsketch_calc(0.5, uddsketch_merge(128, 0.01, udd_state)) as udd_result,
hll_count(hll_merge(hll_state)) as hll_result
FROM
sink_table;
+-+-+
| plan_type_| plan_|
+-+-+
| logical_plan_| Projection: uddsketch_calc(Float64(0.5), uddsketch_merge(Int64(128),Float64(0.01),uddsketch_merge(Int64(128),Float64(0.01),sink_table.udd_state))) AS udd_result, hll_count(hll_merge(hll_merge(sink_table.hll_state))) AS hll_result_|
|_|_Aggregate: groupBy=[[]], aggr=[[uddsketch_merge(Int64(128), Float64(0.01), uddsketch_merge(Int64(128),Float64(0.01),sink_table.udd_state)), hll_merge(hll_merge(sink_table.hll_state))]]_|
|_|_MergeScan [is_placeholder=false, remote_input=[_|
|_| Aggregate: groupBy=[[]], aggr=[[uddsketch_merge(Int64(128), Float64(0.01), sink_table.udd_state), hll_merge(sink_table.hll_state)]]_|
|_|_TableScan: sink_table_|
|_| ]]_|
| physical_plan | ProjectionExec: expr=[uddsketch_calc(0.5, uddsketch_merge(Int64(128),Float64(0.01),uddsketch_merge(Int64(128),Float64(0.01),sink_table.udd_state))@0) as udd_result, hll_count(hll_merge(hll_merge(sink_table.hll_state))@1) as hll_result] |
|_|_AggregateExec: mode=Final, gby=[], aggr=[uddsketch_merge(Int64(128),Float64(0.01),uddsketch_merge(Int64(128),Float64(0.01),sink_table.udd_state)), hll_merge(hll_merge(sink_table.hll_state))]_|
|_|_CoalescePartitionsExec_|
|_|_AggregateExec: mode=Partial, gby=[], aggr=[uddsketch_merge(Int64(128),Float64(0.01),uddsketch_merge(Int64(128),Float64(0.01),sink_table.udd_state)), hll_merge(hll_merge(sink_table.hll_state))]_|
|_|_MergeScanExec: REDACTED
|_|_|
+-+-+
-- SQLNESS REPLACE (metrics.*) REDACTED
-- SQLNESS REPLACE (RoundRobinBatch.*) REDACTED
-- SQLNESS REPLACE (-+) -
-- SQLNESS REPLACE (\s\s+) _
-- SQLNESS REPLACE (peers.*) REDACTED
-- SQLNESS REPLACE region=\d+\(\d+,\s+\d+\) region=REDACTED
-- might write to different partitions
-- SQLNESS REPLACE "partition_count":\{(.*?)\} "partition_count":REDACTED
-- SQLNESS REPLACE (Hash.*) REDACTED
EXPLAIN ANALYZE
SELECT
uddsketch_calc(0.5, uddsketch_merge(128, 0.01, udd_state)) as udd_result,
hll_count(hll_merge(hll_state)) as hll_result
FROM
sink_table;
+-+-+-+
| stage | node | plan_|
+-+-+-+
| 0_| 0_|_ProjectionExec: expr=[uddsketch_calc(0.5, uddsketch_merge(Int64(128),Float64(0.01),uddsketch_merge(Int64(128),Float64(0.01),sink_table.udd_state))@0) as udd_result, hll_count(hll_merge(hll_merge(sink_table.hll_state))@1) as hll_result] REDACTED
|_|_|_AggregateExec: mode=Final, gby=[], aggr=[uddsketch_merge(Int64(128),Float64(0.01),uddsketch_merge(Int64(128),Float64(0.01),sink_table.udd_state)), hll_merge(hll_merge(sink_table.hll_state))] REDACTED
|_|_|_CoalescePartitionsExec REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[], aggr=[uddsketch_merge(Int64(128),Float64(0.01),uddsketch_merge(Int64(128),Float64(0.01),sink_table.udd_state)), hll_merge(hll_merge(sink_table.hll_state))] REDACTED
|_|_|_MergeScanExec: REDACTED
|_|_|_|
| 1_| 0_|_AggregateExec: mode=Final, gby=[], aggr=[uddsketch_merge(Int64(128),Float64(0.01),sink_table.udd_state), hll_merge(sink_table.hll_state)] REDACTED
|_|_|_CoalescePartitionsExec REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[], aggr=[uddsketch_merge(Int64(128),Float64(0.01),sink_table.udd_state), hll_merge(sink_table.hll_state)] REDACTED
|_|_|_SeqScan: region=REDACTED, "partition_count":REDACTED REDACTED
|_|_|_|
| 1_| 1_|_AggregateExec: mode=Final, gby=[], aggr=[uddsketch_merge(Int64(128),Float64(0.01),sink_table.udd_state), hll_merge(sink_table.hll_state)] REDACTED
|_|_|_CoalescePartitionsExec REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[], aggr=[uddsketch_merge(Int64(128),Float64(0.01),sink_table.udd_state), hll_merge(sink_table.hll_state)] REDACTED
|_|_|_SeqScan: region=REDACTED, "partition_count":REDACTED REDACTED
|_|_|_|
| 1_| 2_|_AggregateExec: mode=Final, gby=[], aggr=[uddsketch_merge(Int64(128),Float64(0.01),sink_table.udd_state), hll_merge(sink_table.hll_state)] REDACTED
|_|_|_CoalescePartitionsExec REDACTED
|_|_|_AggregateExec: mode=Partial, gby=[], aggr=[uddsketch_merge(Int64(128),Float64(0.01),sink_table.udd_state), hll_merge(sink_table.hll_state)] REDACTED
|_|_|_SeqScan: region=REDACTED, "partition_count":REDACTED REDACTED
|_|_|_|
|_|_| Total rows: 1_|
+-+-+-+
DROP TABLE integers;
Affected Rows: 0
DROP TABLE sink_table;
Affected Rows: 0