Question

我的查询是他们使用100％CPU。我试图解决它们，但由于我不是数据库专家，它真的变得更糟。

首先我的脚本执行此查询。

SELECT *
FROM 
    (SELECT
       [e].[event_id], 
       [e].[event_name],
       [e].[event_datetime],
       [v].[name] [venue_name],
       [v].[city],
       [s].[state_name],
       [v].[venue_id],
       ROW_NUMBER() OVER (ORDER BY [event_datetime]) AS RowNum
    FROM 
       [indux].[dbo].[event] [e]
    JOIN 
       [indux].[dbo].[venue] [v] ON [e].[venue_id] = [v].[venue_id]
    JOIN 
       [indux].[dbo].[system_state] [s] ON [v].[state_id] = [s].[state_id]
    WHERE 
       [e].[event_id] IN (SELECT DISTINCT event_id 
                          FROM [indux].[dbo].ticket_group 
                          WHERE ticket_group_id IN (SELECT DISTINCT ticket_group_id 
                                                    FROM [indux].[dbo].[ticket] [t] 
                                                    WHERE [t].[actual_sold_price] = -1)
                         )
       AND [e].[event_datetime] >= '2014/08/09'
       AND [e].[event_datetime] <= '2014/12/09'
) AS MyDerivedTable
WHERE 
   MyDerivedTable.RowNum BETWEEN 0 AND 250

并且在同一时间我执行以下查询，而不是将两者的结果结合起来使用PHP在我的页面中显示

SELECT
   COUNT([e].[event_id])
FROM 
   [indux].[dbo].[event] [e]
WHERE 
   [e].[event_id] IN (SELECT DISTINCT event_id 
                      FROM [indux].[dbo].ticket_group 
                      WHERE ticket_group_id IN
                                    (SELECT DISTINCT ticket_group_id 
                                     FROM [indux].[dbo].[ticket] [t] 
                                     WHERE [t].[actual_sold_price] = -1)
                     )
   AND [e].[event_datetime] >= '$today'
   AND [e].[event_datetime] <= '$to'

任何人都知道我是否可以修复这些查询？我认为第二个查询中的计数是个问题。

Answer 1

您需要在故障单组ID上加入故障单组和故障单，并将其加入到事件ID的主查询中。

嵌套查询对于嵌套了where子句的查询的每一行都运行一次。

嵌套两次就像在循环中运行循环一样，整个事情必须运行一次，并且每次查询返回一行时都要将值与值进行比较。从您的选择中可以明显看出，您可以加入这些价值观。

SELECT  *Whatever* 
FROM [indux].[dbo].[event] [e]
           JOIN [indux].[dbo].[venue] [v]
        ON [e].[venue_id] = [v].[venue_id]
      JOIN [indux].[dbo].[system_state] [s]
      ON [v].[state_id] = [s].[state_id]
JOIN indux.dbo.ticket_group tg on tg.event_id=r.event_id
JOIN  indux.dbo.ticket t on t.ticket_group_id=tg.ticket_group_id
WHERE t.actual_sold_price=-1

SELECT COUNT(e.event_id)
FROM dbo.event e
JOIN indux.dbo.ticket_group tg on tg.event_id=e.event_id
JOIN  indux.dbo.ticket t on t.ticket_group_id=tg.ticket_group_id
WHERE e.event_datetime >= '$today' 
AND e.event_datetime <= '$to'

编辑清晰度

SELECT  *Whatever* 
FROM [indux].[dbo].[event] [e]
           JOIN [indux].[dbo].[venue] [v]
        ON [e].[venue_id] = [v].[venue_id]
      JOIN [indux].[dbo].[system_state] [s]
      ON [v].[state_id] = [s].[state_id]
JOIN (
      SELECT DISTINCT tg.event_id 
      FROM indux.dbo.ticket_group tg 
      JOIN indux.dbo.ticket t on t.ticket_group_id=tg.ticket_group_id
      WHERE actual_sold_price = -1) de
on de.Event_id=e.event_id

使用临时表即

CREATE TABLE #EventID (id int) 
INSERT INTO #EventID SELECT tg.event_id etc...

将允许您重用表而不是重新查询count语句。

SQL Server Query占用100％的CPU使用率我的查询有什么问题？

1 个答案: