Details
-
Bug
-
Status: Open (View Workflow)
-
Critical
-
Resolution: Unresolved
-
10.11.15
-
None
-
None
Description
You can simply reproduce with the internal query :
SELECT * FROM `sys` . `statements_with_temp_tables`;
|
The number of SUM_ROWS_SENT is not accurate and mess up with my monitoring : 6 148 914 691 236 517 205
MariaDB [pmacontrol]> select * from performance_schema.events_statements_summary_by_digest where sum_rows_sent > 500000000\G |
*************************** 1. row *************************** |
SCHEMA_NAME: pmacontrol
|
DIGEST: fe37cd99539625fc7da88356f6bb891e
|
DIGEST_TEXT: SELECT * FROM `sys` . `statements_with_temp_tables`
|
COUNT_STAR: 1 |
SUM_TIMER_WAIT: 1600861000 |
MIN_TIMER_WAIT: 1600861000 |
AVG_TIMER_WAIT: 1600861000 |
MAX_TIMER_WAIT: 1600861000 |
SUM_LOCK_TIME: 169000000 |
SUM_ERRORS: 0 |
SUM_WARNINGS: 0 |
SUM_ROWS_AFFECTED: 0 |
SUM_ROWS_SENT: 6148914691236517205 |
SUM_ROWS_EXAMINED: 168 |
SUM_CREATED_TMP_DISK_TABLES: 0 |
SUM_CREATED_TMP_TABLES: 0 |
SUM_SELECT_FULL_JOIN: 0 |
SUM_SELECT_FULL_RANGE_JOIN: 0 |
SUM_SELECT_RANGE: 0 |
SUM_SELECT_RANGE_CHECK: 0 |
SUM_SELECT_SCAN: 1 |
SUM_SORT_MERGE_PASSES: 0 |
SUM_SORT_RANGE: 0 |
SUM_SORT_ROWS: 33 |
SUM_SORT_SCAN: 1 |
SUM_NO_INDEX_USED: 1 |
SUM_NO_GOOD_INDEX_USED: 0 |
FIRST_SEEN: 2025-11-16 15:54:07 |
LAST_SEEN: 2025-11-16 15:54:07 |
1 row in set (0,001 sec) |
After I opened the view to know exactly where we go the problem :
you can simply reproduce with :
SELECT `sys` . `format_time` ( `performance_schema` . `events_statements_summary_by_digest` . `SUM_TIMER_WAIT` ) AS `total_latency` FROM `performance_schema` . `events_statements_summary_by_digest`
|
There is an overflow somewhere, I know it's old but could be really cool to fix it.
