Question

我正在使用以下查询来检索一组订单：

select count(distinct po.orderid) 
from PostOrders po, 
     ProcessedOrders pro 
where pro.state IN ('PENDING','COMPLETED') 
  and po.comp_code in (3,4) 
  and pro.orderid = po.orderid

查询返回结果4323，并且速度足够快。

但是我必须设置另一个条件，只有当它不存在于另一个表DiscarderOrders中时才会返回，我为查询添加了一个额外的条件：

select count(distinct po.orderid) 
from PostOrders po, 
     ProcessedOrders pro 
where pro.state IN ('PENDING','COMPLETED') 
  and po.comp_code in (3,4) 
  and pro.orderid = po.orderid 
  and po.orderid not in (select do.order_id from DiscardedOrders do)

上述查询需要花费大量时间才能继续运行。我可以对查询做些什么，以便快速执行？或者我是否需要先执行第一个查询，然后通过拍摄另一个查询来根据条件进行过滤？

Answer 1

您可以尝试替换：

and po.orderid not in (select do.order_id from DiscardedOrders do)

通过

and not exists (select 1 from DiscardedOrders do where do.order_id = po.orderid)

Answer 2

尝试使用JOIN而不是NOT IN

SELECT  COUNT(DISTINCT po.orderid) TotalCount
FROM    PostOrders po
        INNER JOIN ProcessedOrders pro
            ON po.orderid = pro.orderid
        LEFT JOIN DiscardedOrders do
            ON po.orderid = do.orderid
WHERE   po.comp_code IN (3,4) AND
        pro.state IN ('PENDING','COMPLETED') AND
        do.orderid IS NULL

或NOT EXISTS

SELECT  COUNT(DISTINCT po.orderid) TotalCount
FROM    PostOrders po
        INNER JOIN ProcessedOrders pro
            ON po.orderid = pro.orderid
WHERE   po.comp_code IN (3,4) AND
        pro.state IN ('PENDING','COMPLETED') AND
        NOT EXISTS
        (
            SELECT  1
            FROM    DiscardedOrders do
            WHERE   po.orderid = do.orderid
        )

Answer 3

与distinct相同：

SELECT count(po.orderid) 
  FROM PostOrders po 
 WHERE po.comp_code IN (3,4) 
   AND EXISTS 
   (
   SELECT 1 FROM ProcessedOrders pro 
    WHERE pro.orderid = po.orderid
      AND pro.state IN ('PENDING','COMPLETED') 
   )
  ....
 /

SQL查询优化：花费大量时间

3 个答案: