thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-11-22 15:42:02 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #18763 from azat/marks-corruption-test

Browse Source 
Add a regression test for marks corruption
This commit is contained in:
alexey-milovidov 2021-01-06 21:20:55 +03:00 committed by GitHub
commit 49f70b3b5f
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 61 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
tests/queries/0_stateless/01640_marks_corruption_regression.reference Normal file
View File

@ -0,0 +1,4 @@
"marks",7
"optimize_trivial_count_query",16384
"max_threads=1",16384
"max_threads=100",16384

57
tests/queries/0_stateless/01640_marks_corruption_regression.sql Normal file
View File

@ -0,0 +1,57 @@
DROP TABLE IF EXISTS adaptive_table;
CREATE TABLE adaptive_table(
key UInt64,
value String
) ENGINE MergeTree()
ORDER BY key
SETTINGS
index_granularity_bytes=1048576,
min_bytes_for_wide_part=0,
old_parts_lifetime=0,
index_granularity=8192
;
-- This triggers adjustment of the granules that was introduced in PR#17120
INSERT INTO adaptive_table SELECT number, randomPrintableASCII(if(number BETWEEN 8192-30 AND 8192, 102400, 1)) FROM system.numbers LIMIT 16384;
-- This creates the following marks:
--
-- $ check-marks /path/to/db/adaptive_table/all_*/key.{mrk2,bin}
-- Mark 0, points to 0, 0, has rows after 8192, decompressed size 72808. <!-- wrong number of rows, should be 5461
-- Mark 1, points to 0, 43688, has rows after 1820, decompressed size 29120.
-- Mark 2, points to 0, 58248, has rows after 1820, decompressed size 14560.
-- Mark 3, points to 36441, 0, has rows after 1820, decompressed size 58264.
-- Mark 4, points to 36441, 14560, has rows after 1820, decompressed size 43704.
-- Mark 5, points to 36441, 29120, has rows after 8192, decompressed size 29144.
-- Mark 6, points to 36441, 58264, has rows after 0, decompressed size 0.
OPTIMIZE TABLE adaptive_table FINAL;
SELECT 'marks', marks FROM system.parts WHERE table = 'adaptive_table' AND database = currentDatabase() AND active FORMAT CSV;
-- Reset marks cache
DETACH TABLE adaptive_table;
ATTACH TABLE adaptive_table;
-- This works correctly, since it does not read any marks
SELECT 'optimize_trivial_count_query', count() FROM adaptive_table SETTINGS
optimize_trivial_count_query=1
FORMAT CSV;
-- This works correctly, since it reads marks sequentially and don't seek
SELECT 'max_threads=1', count() FROM adaptive_table SETTINGS
optimize_trivial_count_query=0,
max_threads=1
FORMAT CSV;
-- This works wrong, since it seek to each mark (due to reading each mark from a separate thread),
-- so if the marks offsets will be wrong it will read more data.
--
-- Reading each mark from a separate thread is just the simplest reproducers,
-- this can be also reproduced with PREWHERE since it skips data physically,
-- so it also uses seeks.
SELECT 'max_threads=100', count() FROM adaptive_table SETTINGS
optimize_trivial_count_query=0,
merge_tree_min_rows_for_concurrent_read=1,
merge_tree_min_bytes_for_concurrent_read=1,
max_threads=100
FORMAT CSV;
DROP TABLE adaptive_table;