Skip to content

S3ReadRequestsCount significantly differs (x2) for Iceberg tables between 25.11 and head #93416

@alsugiliazova

Description

@alsugiliazova

Describe what's wrong

I run these queries on head (26.1.1.277):

I have iceberg table partitioned by integer column:

SELECT *
FROM database_c7100729_ea40_11f0_b919_e0c26496f172.`namespace_c71006fa_ea40_11f0_af67_e0c26496f172.table_c71006ad_ea40_11f0_ab4f_e0c26496f172`
ORDER BY integer ASC
FORMAT TabSeparated
0	true	0
1	false	1
2	true	2
3	false	3
4	true	4
5	false	5
6	true	6
7	false	7
8	true	8
9	false	9
10	true	10
11	false	11
12	true	12
13	false	13
14	true	14
15	false	15
16	true	16
17	false	17
18	true	18
19	false	19

I do simple select:

SELECT *
FROM database_e05669b1_ea41_11f0_b919_e0c26496f172.`namespace_e0566978_ea41_11f0_af67_e0c26496f172.table_e056692f_ea41_11f0_ab4f_e0c26496f172`
WHERE integer >= 0
ORDER BY integer ASC
SETTINGS log_comment = 'count_s3_requests'
Query id: 4394b99e-c601-4ae3-8e54-75cb1b0eec03

    ┌─string─┬─boolean─┬─integer─┐
 1. │ 0      │ true    │       0 │
 2. │ 1      │ false   │       1 │
 3. │ 2      │ true    │       2 │
 4. │ 3      │ false   │       3 │
 5. │ 4      │ true    │       4 │
 6. │ 5      │ false   │       5 │
 7. │ 6      │ true    │       6 │
 8. │ 7      │ false   │       7 │
 9. │ 8      │ true    │       8 │
10. │ 9      │ false   │       9 │
11. │ 10     │ true    │      10 │
12. │ 11     │ false   │      11 │
13. │ 12     │ true    │      12 │
14. │ 13     │ false   │      13 │
15. │ 14     │ true    │      14 │
16. │ 15     │ false   │      15 │
17. │ 16     │ true    │      16 │
18. │ 17     │ false   │      17 │
19. │ 18     │ true    │      18 │
20. │ 19     │ false   │      19 │
    └────────┴─────────┴─────────┘

20 rows in set. Elapsed: 0.048 sec. 
SELECT ProfileEvents['S3ReadRequestsCount']
FROM system.query_log
WHERE (log_comment = 'count_s3_requests') AND (type = 'QueryFinish')
Query id: 450eb0e5-382a-45dd-a2f0-c24493fcc7cd

   ┌─arrayElement⋯estsCount')─┐
1. │                       86 │
   └──────────────────────────┘

1 row in set. Elapsed: 0.004 sec. 

I do the same on 25.11:

ClickHouse client version 25.11.5.8 (official build).
Connecting to localhost:9000 as user default.
Connected to ClickHouse server version 25.11.5.

SELECT *
FROM database_5e9dc2c2_ea42_11f0_b919_e0c26496f172.`namespace_5e9dc28a_ea42_11f0_af67_e0c26496f172.table_5e9dc23d_ea42_11f0_ab4f_e0c26496f172`
WHERE integer >= 0
ORDER BY integer ASC
SETTINGS log_comment = 'count_s3_requests'
Query id: 99ccc880-8e51-4eec-84ec-733317a26377

    ┌─string─┬─boolean─┬─integer─┐
 1. │ 0      │ true    │       0 │
 2. │ 1      │ false   │       1 │
 3. │ 2      │ true    │       2 │
 4. │ 3      │ false   │       3 │
 5. │ 4      │ true    │       4 │
 6. │ 5      │ false   │       5 │
 7. │ 6      │ true    │       6 │
 8. │ 7      │ false   │       7 │
 9. │ 8      │ true    │       8 │
10. │ 9      │ false   │       9 │
11. │ 10     │ true    │      10 │
12. │ 11     │ false   │      11 │
13. │ 12     │ true    │      12 │
14. │ 13     │ false   │      13 │
15. │ 14     │ true    │      14 │
16. │ 15     │ false   │      15 │
17. │ 16     │ true    │      16 │
18. │ 17     │ false   │      17 │
19. │ 18     │ true    │      18 │
20. │ 19     │ false   │      19 │
    └────────┴─────────┴─────────┘

20 rows in set. Elapsed: 0.038 sec. 
SELECT ProfileEvents['S3ReadRequestsCount']
FROM system.query_log
WHERE (log_comment = 'count_s3_requests') AND (type = 'QueryFinish')
Query id: 1e2a9bc3-72a2-4474-9e91-a9979721e2fa

   ┌─arrayElement⋯estsCount')─┐
1. │                       44 │
   └──────────────────────────┘

1 row in set. Elapsed: 0.004 sec. 

Does it reproduce on the most recent release?

Yes

Metadata

Metadata

Labels

comp-datalakeData lake table formats (Iceberg/Delta/Hudi) integration.unexpected behaviourResult is unexpected, but not entirely wrong at the same time.

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions