在一次的优化过程中，由于没有关注执行计划中type列，仅看key列来查看"使用到的索引"，导致优化过程走了不少弯路。

以下面SQL为例：

SELECT wave_no,
SUM(IF(picking_qty IS NULL, 0, picking_qty)) AS PICKED_QTY,
SUM(IF(differ_qty IS NULL, 0, differ_qty)) AS PICKED_DIFFER_QTY,
SUM(IF(relocate_qty IS NULL, 0, relocate_qty)) AS PICKED_RELOCATE_QTY 
FROM picking_locate_d
WHERE yn = 0
AND wave_no IN
(
'BC76361213164811',
'BC76361213164810',
...
'BC76361213158692'
)
AND org_No = '661'
AND distribute_No = '763'
AND warehouse_No = '612' 
GROUP BY wave_no;

走索引查找的执行计划为：

+----+-------------+------------------+------------+-------+---------------+-------------+---------+------+-------+----------+------------------------------------+
| id | select_type | table            | partitions | type  | possible_keys | key         | key_len | ref  | rows  | filtered | Extra                              |
+----+-------------+------------------+------------+-------+---------------+-------------+---------+------+-------+----------+------------------------------------+
|  1 | SIMPLE      | picking_locate_d | NULL       | range | idx_wave_no   | idx_wave_no | 153     | NULL | 14238 |     0.01 | Using index condition; Using where |
+----+-------------+------------------+------------+-------+---------------+-------------+---------+------+-------+----------+------------------------------------+