SQL 优化慢的update语句

当归10242024-08-21 16:49

原语句

sql 复制代码

 UPDATE tbl_mix_trace
 SET mix_plan_id = ( SELECT p.id FROM tbl_mix_plan p WHERE p.lot_id_out= lot_id_in ),
 pro_date_in = (SELECT p.pro_date_out FROM tbl_mix_plan p WHERE p.lot_id_out = lot_id_in),
 shift_id_in = (SELECT p.shift_id_out FROM tbl_mix_plan p WHERE p.lot_id_out = lot_id_in),
 equip_id_in = (SELECT p.equip_id_out FROM tbl_mix_plan p WHERE p.lot_id_out = lot_id_in)
 where   <![CDATA[
     ( lot_id_in <> '' and lot_id_in is not NULL)
    ]]>
 and pro_date_out = #{proDateOut}
 and shift_id_out = #{shiftIdOut}

原因

使用了过多的子查询，查询的还是同一个表，效率低

使用join 可以避免这个问题（不同的数据库写法可能不同）

更改后的语句

sql 复制代码

UPDATE t
SET 
    t.mix_plan_id = p.id,
    t.pro_date_in = p.pro_date_out,
    t.shift_id_in = p.shift_id_out,
    t.equip_id_in = p.equip_id_out
FROM 
    tbl_mix_trace t
JOIN 
    tbl_mix_plan p ON p.lot_id_out = t.lot_id_in
WHERE 
    t.lot_id_in <> '' AND t.lot_id_in IS NOT NULL
    AND t.pro_date_out = '2024-08-14'
    AND t.shift_id_out = 1;