MySQL的rows_examined指标缺陷总结

rows_examined是什么？
https://dev.mysql.com/doc/refman/8.4/en/slow-query-log.html
文档上的slow query log部分有解释：
The number of rows examined by the server layer (not counting any processing internal to storage engines).

rows examined实际上是server层检查的行数，没有统计storage engine层内部处理的行数。这样很多看似是有问题的，就很好理解了，

但是，很多人把rows examined当成SQL性能的一种重要指标，认为这个是SQL扫描的行数，这就有问题了，实际上rows examined有时候，并不能准确反应storage engines扫描的行数。

engine层返回给server层数据，engine层对表使用访问方法：索引访问、全表扫描，过滤条件也即执行计划里的Filter是在server层做的。


如果索引访问有ICP，则可能导致rows_examined并不能反应索引真正扫描的行数，还有其它缺陷，比如分组聚合、排序影响。


总结如下：
1）二级索引的ICP，会导致返回给server层行数变少，所以此时的rows_examined是包含ICP的结果，并不能准确反应索引扫描的行数。
   而覆盖索引扫描，是没有ICP的，过滤发生在server层，则是准确的。


2）分组聚合如果没有其他字段会让Rows_examined是0，这应该是个BUG


3）index skip scan貌似是0，禁用skip scan则rows_examined不是0


4）排序的行数会加到Rows_examined里，如果有limit，rows_examined会加上limit行数。

1.有ICP时候rows_examined不能反应索引扫描的行数

explain select * from employees e where e.first_name like 'Am%' and e.last_name like 'B%';
+----+-------------+-------+------------+-------+-------------------+----------+---------+------+------+----------+-----------------------+
| id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra |
+----+-------------+-------+------------+-------+-------------------+----------+---------+------+------+----------+-----------------------+
| 1 | SIMPLE | e | NULL | range | idx1_emp,idx2_emp | idx1_emp | 94 | NULL | 696 | 11.11 | Using index condition |
+----+-------------+-------+------------+-------+-------------------+----------+---------+------+------+----------+-----------------------+
1 row in set, 1 warning (0.00 sec)

flush status;
show status like '%handler_read%';
+-----------------------+-------+
| Variable_name | Value |
+-----------------------+-------+
| Handler_read_first | 0 |
| Handler_read_key | 0 |
| Handler_read_last | 0 |
| Handler_read_next | 0 |
| Handler_read_prev | 0 |
| Handler_read_rnd | 0 |
| Handler_read_rnd_next | 0 |
+-----------------------+-------+
7 rows in set (0.00 sec)

 explain analyze select * from employees e where e.first_name like 'Am%' and e.last_name like 'B%'\G
*************************** 1. row ***************************
EXPLAIN: -> Index range scan on e using idx1_emp over
('Am\0\0\0\0\0\0\0\0\0\0\0\0' <= first_name <= 'Am????????????' AND 'B\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0'
<= last_name <= 'B???????????????'), with index condition: ((e.first_name like 'Am%') and (e.last_name like 'B%'))
 (cost=836.21 rows=696) (actual time=3.599..65.615 rows=69 loops=1)

1 row in set (0.07 sec)

(root@localhost) [employees_10:58:41]> show status like '%handler_read%';
+-----------------------+-------+
| Variable_name | Value |
+-----------------------+-------+
| Handler_read_first | 0 |
| Handler_read_key | 1 |
| Handler_read_last | 0 |
| Handler_read_next | 69 |
| Handler_read_prev | 0 |
| Handler_read_rnd | 0 |
| Handler_read_rnd_next | 0 |
+-----------------------+-------+
7 rows in set (0.00 sec)

select THREAD_ID,SQL_TEXT,ROWS_AFFECTED,ROWS_SENT,ROWS_EXAMINED from performance_schema.events_statements_history where thread_id=48;
+-----------+---------------------------------------------------------------------------------------------------+---------------+-----------+---------------+
| THREAD_ID | SQL_TEXT | ROWS_AFFECTED | ROWS_SENT | ROWS_EXAMINED |
+-----------+---------------------------------------------------------------------------------------------------+---------------+-----------+---------------+
| 48 | show status like '%handler_read%' | 0 | 7 | 7 |
| 48 | explain analyze select * from employees e where e.first_name like 'Am%' and e.last_name like 'B%' | 0 | 1 | 69 |
+-----------+---------------------------------------------------------------------------------------------------+---------------+-----------+---------------+
10 rows in set (0.00 sec)

# Time: 2025-03-14T10:58:41.069062+08:00
# User@Host: root[root] @ localhost [] Id: 8
# Query_time: 0.066471  Lock_time: 0.000005 Rows_sent: 1 Rows_examined: 69 Thread_id: 8 Errno: 0 Killed: 0 Bytes_received: 104 Bytes_sent: 440 Read_first: 0 Read_last: 0 Read_key: 1 Read_next: 69 Read_prev: 0 Read_rnd: 0 Read_rnd_next: 0 Sort_merge_passes: 0 Sort_range_count: 0 Sort_rows: 0 Sort_scan_count: 0 Created_tmp_disk_tables: 0 Created_tmp_tables: 0 Start: 2025-03-14T10:58:41.002591+08:00 End: 2025-03-14T10:58:41.069062+08:00
SET timestamp=1741921121;
explain analyze select * from employees e where e.first_name like 'Am%' and e.last_name like 'B%';

实际行数：实际上index access只按前导列，应该是扫描696行，但是rows_examined=69行，这是MySQL缺陷，
MySQL的rows_examined实际上是engine返回给server层的行数，因为有ICP，导致返回给server层行数包括ICP，
对应的行数是engine(index)=69行--->engine(table)回表 69行--->server 69行：

select count(*) from employees e where e.first_name like 'Am%';
+----------+
| count(*) |
+----------+
| 696 |
+----------+

 select count(*) from employees e where e.first_name like 'Am%' and e.last_name like 'B%' ;
+----------+
| count(*) |
+----------+
| 69 |
+----------+

 如果没有ICP，比如下面的是覆盖索引扫描，则是696行，直接访问表扫描696行，过滤是在server层做的。

explain select first_name,last_name from employees e where e.first_name like 'Am%' and e.last_name like 'B%'
    -> ;
+----+-------------+-------+------------+-------+-------------------+----------+---------+------+------+----------+--------------------------+
| id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra |
+----+-------------+-------+------------+-------+-------------------+----------+---------+------+------+----------+--------------------------+
| 1 | SIMPLE | e | NULL | range | idx1_emp,idx2_emp | idx1_emp | 94 | NULL | 696 | 11.11 | Using where; Using index |
+----+-------------+-------+------------+-------+-------------------+----------+---------+------+------+----------+--------------------------+
1 row in set, 1 warning (0.00 sec)

explain analyze select first_name,last_name from employees e where e.first_name like 'Am%' and e.last_name like 'B%'\G
*************************** 1. row ***************************
EXPLAIN: -> Filter: ((e.first_name like 'Am%') and (e.last_name like 'B%')) (cost=141.53 rows=77) (actual time=0.053..0.584 rows=69 loops=1)
    -> Covering index range scan on e using idx1_emp over ('Am\0\0\0\0\0\0\0\0\0\0\0\0' <= first_name <= 'Am????????????' AND 'B\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0' <= last_name <= 'B???????????????') (cost=141.53 rows=696) (actual time=0.030..0.431 rows=696 loops=1)

1 row in set (0.00 sec)

# Time: 2025-03-14T14:31:44.984084+08:00
# User@Host: root[root] @ localhost [] Id: 15
# Query_time: 0.001385  Lock_time: 0.000004 Rows_sent: 1 Rows_examined: 696 Thread_id: 15 Errno: 0 Killed: 0 Bytes_received: 123 Bytes_sent: 507 Read_first: 0 Read_last: 0 Read_key: 1 Read_next: 696 Read_prev: 0 Read_rnd: 0 Read_rnd_next: 0 Sort_merge_passes: 0 Sort_range_count: 0 Sort_rows: 0 Sort_scan_count: 0 Created_tmp_disk_tables: 0 Created_tmp_tables: 0 Start: 2025-03-14T14:31:44.982699+08:00 End: 2025-03-14T14:31:44.984084+08:00
SET timestamp=1741933904;
explain analyze select first_name,last_name from employees e where e.first_name like 'Am%' and e.last_name like 'B%';

可以看到，覆盖索引扫描的Rows_examined: 696。


2.分组聚合,如果是通过索引直接聚合的会让Rows_examined是0，如果是全表扫描则没有问题

无条件的聚合操作，Rows_examined: 0，因为走索引

select count(*) from employees e;
+----------+
| count(*) |
+----------+
| 300024 |
+----------+
1 row in set (0.14 sec)

explain select count(*) from employees e;
+----+-------------+-------+------------+-------+---------------+----------+---------+------+--------+----------+-------------+
| id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra |
+----+-------------+-------+------------+-------+---------------+----------+---------+------+--------+----------+-------------+
| 1 | SIMPLE | e | NULL | index | NULL | idx3_emp | 1 | NULL | 300024 | 100.00 | Using index |
+----+-------------+-------+------------+-------+---------------+----------+---------+------+--------+----------+-------------+
1 row in set, 1 warning (0.01 sec)

select THREAD_ID,SQL_TEXT,ROWS_AFFECTED,ROWS_SENT,ROWS_EXAMINED from performance_schema.events_statements_history where thread_id=48;
+-----------+---------------------------------------------------------------------------------------------------+---------------+-----------+---------------+
| THREAD_ID | SQL_TEXT | ROWS_AFFECTED | ROWS_SENT | ROWS_EXAMINED |
+-----------+---------------------------------------------------------------------------------------------------+---------------+-----------+---------------+
| 48 | select count(*) from employees e | 0 | 1 | 0 |


# Time: 2025-03-14T11:02:53.082873+08:00
# User@Host: root[root] @ localhost [] Id: 8
# Query_time: 0.130915  Lock_time: 0.000005 Rows_sent: 1  Rows_examined: 0 Thread_id: 8 Errno: 0 Killed: 0 Bytes_received: 39 Bytes_sent: 61 Read_first: 0 Read_last: 0 Read_key: 0 Read_next: 0 Read_prev: 0 Read_rnd: 0 Read_rnd_next: 0 Sort_merge_passes: 0 Sort_range_count: 0 Sort_rows: 0 Sort_scan_count: 0 Created_tmp_disk_tables: 0 Created_tmp_tables: 0 Start: 2025-03-14T11:02:52.951958+08:00 End: 2025-03-14T11:02:53.082873+08:00
SET timestamp=1741921372;
select count(*) from employees e;


有条件的也是Rows_examined: 0，因为走last_name列索引
select count(*) from employees e where last_name like 'A%';
+----------+
| count(*) |
+----------+
| 10673 |
+----------+


# Time: 2025-03-14T11:03:51.564739+08:00
# User@Host: root[root] @ localhost [] Id: 8
# Query_time: 0.107009  Lock_time: 0.000005 Rows_sent: 1  Rows_examined: 0 Thread_id: 8 Errno: 0 Killed: 0 Bytes_received: 65 Bytes_sent: 60 Read_first: 1 Read_last: 0 Read_key: 2551 Read_next: 10673 Read_prev: 0 Read_rnd: 0 Read_rnd_next: 0 Sort_merge_passes: 0 Sort_range_count: 0 Sort_rows: 0 Sort_scan_count: 0 Created_tmp_disk_tables: 0 Created_tmp_tables: 0 Start: 2025-03-14T11:03:51.457730+08:00 End: 2025-03-14T11:03:51.564739+08:00
SET timestamp=1741921431;
select count(*) from employees e where last_name like 'A%';


# Time: 2025-03-14T11:04:47.356855+08:00
# User@Host: root[root] @ localhost [] Id: 8
# Query_time: 0.026070  Lock_time: 0.000005 Rows_sent: 1  Rows_examined: 0 Thread_id: 8 Errno: 0 Killed: 0 Bytes_received: 71 Bytes_sent: 66 Read_first: 1 Read_last: 0 Read_key: 2551 Read_next: 10673 Read_prev: 0 Read_rnd: 0 Read_rnd_next: 0 Sort_merge_passes: 0 Sort_range_count: 0 Sort_rows: 0 Sort_scan_count: 0 Created_tmp_disk_tables: 0 Created_tmp_tables: 0 Start: 2025-03-14T11:04:47.330785+08:00 End: 2025-03-14T11:04:47.356855+08:00
SET timestamp=1741921487;
select max(last_name) from employees e where last_name like 'A%';

-- 全表扫描则没有问题
SELECT/*+no_index(e)*/ COUNT(*) FROM employees e WHERE last_name LIKE 'A%';
+----------+
| COUNT(*) |
+----------+
| 10673 |
+----------+
1 row in set (0.10 sec)

# Time: 2025-03-14T15:06:13.796973+08:00
# User@Host: root[root] @ localhost [] Id: 15
# Query_time: 0.092579  Lock_time: 0.000006 Rows_sent: 1  Rows_examined: 300024 Thread_id: 15 Errno: 0 Killed: 0 Bytes_received: 81 Bytes_sent: 60 Read_first: 1 Read_last: 0 Read_key: 1 Read_next: 0 Read_prev: 0 Read_rnd: 0 Read_rnd_next: 300025 Sort_merge_passes: 0 Sort_range_count: 0 Sort_rows: 0 Sort_scan_count: 0 Created_tmp_disk_tables: 0 Created_tmp_tables: 0 Start: 2025-03-14T15:06:13.704394+08:00 End: 2025-03-14T15:06:13.796973+08:00
SET timestamp=1741935973;
SELECT/*+no_index(e)*/ COUNT(*) FROM employees e WHERE last_name LIKE 'A%';

3.index skip scan貌似是0，禁用skip scan则rows_examined不是0

explain select last_name from employees e where last_name like 'A%';
+----+-------------+-------+------------+-------+-------------------+----------+---------+------+-------+----------+----------------------------------------+
| id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra |
+----+-------------+-------+------------+-------+-------------------+----------+---------+------+-------+----------+----------------------------------------+
| 1 | SIMPLE | e | NULL | range | idx1_emp,idx2_emp | idx1_emp | 94 | NULL | 33332 | 100.00 | Using where; Using index for skip scan |
+----+-------------+-------+------------+-------+-------------------+----------+---------+------+-------+----------+----------------------------------------+
1 row in set, 1 warning (0.00 sec)

-- index skip scan的Rows_examined: 0
# User@Host: root[root] @ localhost [] Id: 8
# Query_time: 0.030014  Lock_time: 0.000006 Rows_sent: 10673  Rows_examined: 0 Thread_id: 8 Errno: 0 Killed: 0 Bytes_received: 66 Bytes_sent: 129296 Read_first: 1 Read_last: 0 Read_key: 2551 Read_next: 10673 Read_prev: 0 Read_rnd: 0 Read_rnd_next: 0 Sort_merge_passes: 0 Sort_range_count: 0 Sort_rows: 0 Sort_scan_count: 0 Created_tmp_disk_tables: 0 Created_tmp_tables: 0 Start: 2025-03-14T11:05:08.350082+08:00 End: 2025-03-14T11:05:08.380096+08:00
SET timestamp=1741921508;
select last_name from employees e where last_name like 'A%';


-- 禁用skip scan则rows_examined不是0
select/*+set_var(optimizer_switch='skip_scan=off')*/ last_name from employees e where last_name like 'A%';

# Time: 2025-03-14T15:18:20.895184+08:00
# User@Host: root[root] @ localhost [] Id: 15
# Query_time: 0.083965  Lock_time: 0.000005 Rows_sent: 10673  Rows_examined: 300024 Thread_id: 15 Errno: 0 Killed: 0 Bytes_received: 112 Bytes_sent: 129296 Read_first: 1 Read_last: 0 Read_key: 1 Read_next: 300024 Read_prev: 0 Read_rnd: 0 Read_rnd_next: 0 Sort_merge_passes: 0 Sort_range_count: 0 Sort_rows: 0 Sort_scan_count: 0 Created_tmp_disk_tables: 0 Created_tmp_tables: 0 Start: 2025-03-14T15:18:20.811219+08:00 End: 2025-03-14T15:18:20.895184+08:00
SET timestamp=1741936700;
select/*+set_var(optimizer_switch='skip_scan=off')*/ last_name from employees e where last_name like 'A%';




有其他字段则可以
select gender,count(*) from employees e where last_name like 'A%' group by gender;
# Time: 2025-03-14T14:49:36.060981+08:00
# User@Host: root[root] @ localhost [] Id: 15
# Query_time: 0.886257  Lock_time: 0.000008 Rows_sent: 2  Rows_examined: 300024 Thread_id: 15 Errno: 0 Killed: 0 Bytes_received: 88 Bytes_sent: 129 Read_first: 1 Read_last: 0 Read_key: 1 Read_next: 300024 Read_prev: 0 Read_rnd: 0 Read_rnd_next: 0 Sort_merge_passes: 0 Sort_range_count: 0 Sort_rows: 0 Sort_scan_count: 0 Created_tmp_disk_tables: 0 Created_tmp_tables: 0 Start: 2025-03-14T14:49:35.174724+08:00 End: 2025-03-14T14:49:36.060981+08:00
SET timestamp=1741934975;
select gender,count(*) from employees e where last_name like 'A%' group by gender;

-- 查点其他字段才对，Rows_examined: 300024

# Time: 2025-03-14T11:05:51.264599+08:00
# User@Host: root[root] @ localhost [] Id: 8
# Query_time: 0.153229  Lock_time: 0.000004 Rows_sent: 1  Rows_examined: 300024 Thread_id: 8 Errno: 0 Killed: 0 Bytes_received: 75 Bytes_sent: 266 Read_first: 1 Read_last: 0 Read_key: 1 Read_next: 0 Read_prev: 0 Read_rnd: 0 Read_rnd_next: 300025 Sort_merge_passes: 0 Sort_range_count: 0 Sort_rows: 0 Sort_scan_count: 0 Created_tmp_disk_tables: 0 Created_tmp_tables: 0 Start: 2025-03-14T11:05:51.111370+08:00 End: 2025-03-14T11:05:51.264599+08:00
SET timestamp=1741921551;
explain analyze select * from  employees e where last_name like 'A%';

4.order by的行会累加到rows_examined,这个应该也是BUG，order by实际上是在server层发生的

-- order by竟然让Rows_examined翻倍
# Time: 2025-03-14T11:06:59.889233+08:00
# User@Host: root[root] @ localhost [] Id: 8
# Query_time: 0.000366  Lock_time: 0.000005 Rows_sent: 10  Rows_examined: 10 Thread_id: 8 Errno: 0 Killed: 0 Bytes_received: 29 Bytes_sent: 131 Read_first: 1 Read_last: 0 Read_key: 1 Read_next: 0 Read_prev: 0 Read_rnd: 0 Read_rnd_next: 11 Sort_merge_passes: 0 Sort_range_count: 0 Sort_rows: 0 Sort_scan_count: 0 Created_tmp_disk_tables: 0 Created_tmp_tables: 0 Start: 2025-03-14T11:06:59.888867+08:00 End: 2025-03-14T11:06:59.889233+08:00
SET timestamp=1741921619;
select * from limit_t1;
# Time: 2025-03-14T11:07:14.836767+08:00
# User@Host: root[root] @ localhost [] Id: 8
# Query_time: 0.000443  Lock_time: 0.000005 Rows_sent: 10  Rows_examined: 20 Thread_id: 8 Errno: 0 Killed: 0 Bytes_received: 41 Bytes_sent: 131 Read_first: 1 Read_last: 0 Read_key: 1 Read_next: 0 Read_prev: 0 Read_rnd: 0 Read_rnd_next: 11 Sort_merge_passes: 0 Sort_range_count: 0 Sort_rows: 10 Sort_scan_count: 1 Created_tmp_disk_tables: 0 Created_tmp_tables: 0 Start: 2025-03-14T11:07:14.836324+08:00 End: 2025-03-14T11:07:14.836767+08:00
SET timestamp=1741921634;
select * from limit_t1 order by id;


-- 走索引，查询出4行，对
# Time: 2025-03-14T11:09:16.819234+08:00
# User@Host: root[root] @ localhost [] Id: 8
# Query_time: 0.000482  Lock_time: 0.000005 Rows_sent: 4  Rows_examined: 4 Thread_id: 8 Errno: 0 Killed: 0 Bytes_received: 52 Bytes_sent: 95 Read_first: 0 Read_last: 0 Read_key: 1 Read_next: 4 Read_prev: 0 Read_rnd: 0 Read_rnd_next: 0 Sort_merge_passes: 0 Sort_range_count: 0 Sort_rows: 0 Sort_scan_count: 0 Created_tmp_disk_tables: 0 Created_tmp_tables: 0 Start: 2025-03-14T11:09:16.818752+08:00 End: 2025-03-14T11:09:16.819234+08:00
SET timestamp=1741921756;
select * from limit_t1 where id>3 order by id;




show create table t\G
*************************** 1. row ***************************
       Table: t
Create Table: CREATE TABLE `t` (
  `id` int NOT NULL,
  `a` int DEFAULT NULL,
  `b` int DEFAULT NULL,
  PRIMARY KEY (`id`),
  KEY `idx_a` (`a`),
  KEY `idx1_t` (`b`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_0900_ai_ci
1 row in set (0.00 sec)

(root@localhost) [employees_11:10:17]> select * from t;
+----+------+------+
| id | a | b |
+----+------+------+
| 1 | 1 | 1 |
| 2 | 1 | 2 |
| 3 | 1 | 3 |
| 4 | 1 | 4 |
| 5 | 1 | 5 |
| 6 | 1 | 6 |
| 7 | 1 | 7 |
| 8 | 1 | 8 |
| 9 | 1 | 9 |
| 10 | 1 | 10 |
+----+------+------+
10 rows in set (0.00 sec)

(root@localhost) [employees_11:10:21]> select * from t where a=1 and b>3 order by b limit 2;
+----+------+------+
| id | a | b |
+----+------+------+
| 4 | 1 | 4 |
| 5 | 1 | 5 |
+----+------+------+
2 rows in set (0.00 sec)

表有10行，order by limit 2,Rows_examined: 12,还是order by问题，因为a列有索引，全是1，读取10行+排序2行=12
# Time: 2025-03-14T11:10:35.769475+08:00
# User@Host: root[root] @ localhost [] Id: 8
# Query_time: 0.000579  Lock_time: 0.000007 Rows_sent: 2 Rows_examined: 12 Thread_id: 8 Errno: 0 Killed: 0 Bytes_received: 59 Bytes_sent: 155 Read_first: 0 Read_last: 0 Read_key: 1 Read_next: 10 Read_prev: 0 Read_rnd: 0 Read_rnd_next: 0 Sort_merge_passes: 0 Sort_range_count: 0 Sort_rows: 2 Sort_scan_count: 1 Created_tmp_disk_tables: 0 Created_tmp_tables: 0 Start: 2025-03-14T11:10:35.768896+08:00 End: 2025-03-14T11:10:35.769475+08:00
SET timestamp=1741921835;
select * from t where a=1 and b>3 order by b limit 2;



explain select * from t where a=1 and b>3 order by b limit 2;
+----+-------------+-------+------------+------+---------------+-------+---------+-------+------+----------+-----------------------------+
| id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra |
+----+-------------+-------+------------+------+---------------+-------+---------+-------+------+----------+-----------------------------+
| 1 | SIMPLE | t | NULL | ref | idx_a,idx1_t | idx_a | 5 | const | 10 | 70.00 | Using where; Using filesort |
+----+-------------+-------+------------+------+---------------+-------+---------+-------+------+----------+-----------------------------+
1 row in set, 1 warning (0.00 sec)

-- 不走索引，也是12行
select/*+no_index(t)*/ * from t where a=1 and b>3 order by b limit 2;

# Time: 2025-03-14T14:54:34.514207+08:00
# User@Host: root[root] @ localhost [] Id: 15
# Query_time: 0.000935  Lock_time: 0.000008 Rows_sent: 2 Rows_examined: 12 Thread_id: 15 Errno: 0 Killed: 0 Bytes_received: 75 Bytes_sent: 155 Read_first: 1 Read_last: 0 Read_key: 1 Read_next: 0 Read_prev: 0 Read_rnd: 0 Read_rnd_next: 11 Sort_merge_passes: 0 Sort_range_count: 0 Sort_rows: 2 Sort_scan_count: 1 Created_tmp_disk_tables: 0 Created_tmp_tables: 0 Start: 2025-03-14T14:54:34.513272+08:00 End: 2025-03-14T14:54:34.514207+08:00
SET timestamp=1741935274;
select/*+no_index(t)*/ * from t where a=1 and b>3 order by b limit 2;