在派生表中不接受選項（散列組）

Question

可以在第一個但不是第二個中使用OPTION（HASH GROUP）。我如何在第二個包含OPTION（HASH GROUP）？

SELECT 
    count(*) as 'match' 
    FROM [docSVenum1] with (nolock) 
    INNER LOOP JOIN [FTSindexWordOnce] as w1 with (NOLOCK, FORCESEEK) 
    ON [docSVenum1].sID = w1.[sID] and [docSVenum1].[enumID] = '142' 
    INNER HASH JOIN [FTSindexWordOnce] as w2 with (NOLOCK) 
    ON w1.wordID = w2.wordID and w2.[sID] = '2'  
    GROUP BY W1.[sID] 
    OPTION (HASH GROUP) 


select max(list.match) as 'max' 
    from 
    (
     SELECT 
     count(*) as 'match' 
     FROM [docSVenum1] with (nolock) 
     INNER LOOP JOIN [FTSindexWordOnce] as w1 with (NOLOCK, FORCESEEK) 
     ON [docSVenum1].sID = w1.[sID] and [docSVenum1].[enumID] = '142' 
     INNER HASH JOIN [FTSindexWordOnce] as w2 with (NOLOCK) 
     ON w1.wordID = w2.wordID and w2.[sID] = '2'  
     GROUP BY W1.[sID] 
     -- OPTION (HASH GROUP) 
) as list; 

在現實生活中充滿查詢會更喜歡這個

select max(list.match) as 'max' 
    from 
    (
     SELECT 200*count(*)/([d1].[textSize] + [d2].[textSize]) as 'match' 
     FROM [docSVenum1] with (nolock) 
     INNER LOOP JOIN [FTSindexWordOnce] as w1 with (NOLOCK, FORCESEEK) 
     ON [docSVenum1].sID = w1.[sID] and [docSVenum1].[enumID] = '142' 
     INNER HASH JOIN [FTSindexWordOnce] as w2 with (NOLOCK) 
     ON w1.wordID = w2.wordID and w2.[sID] = '2' 
     JOIN docSVsys as d1 with (nolock) 
     on d1.sID = w1.sID 
     JOIN docSVsys as d2 with (nolock) 
     on d2.sID = w1.sID  
     GROUP BY W1.[sID], [d1].[textSize], [d2].[textSize] 
     -- OPTION (HASH GROUP) 
) as list; 

Answer 1

我能夠到SQL Server 2005上使用子查詢之外的選項（as list）之後（不含FORCESEEK提示，版本= 9.0.5266），2008（10.0.5785）和2012（11.0.2316）。但在2008 R2（10.50.2811）錯誤消息我得到的是：

消息8622，級別16，狀態1，行1
查詢處理器未能生成因爲在這個定義中的提示的查詢計劃查詢。重新提交查詢而不指定任何提示，也不使用SET FORCEPLAN。

SELECT MAX(x.match) FROM (
SELECT 
    count(*) as match 
    FROM sys.objects AS o with (nolock, forceseek) 
    INNER LOOP JOIN sys.columns as c with (NOLOCK) 
    ON o.object_id = c.object_id 
    INNER HASH JOIN sys.columns as c2 with (NOLOCK) 
    ON o.object_id = c2.object_id 
    GROUP BY o.object_id 
) AS x 
    OPTION (HASH GROUP); 

如果我不使用聚合，它的工作原理：

SELECT x.match FROM (
SELECT 
    count(*) as match 
    FROM sys.objects AS o with (nolock, forceseek) 
    INNER LOOP JOIN sys.columns as c with (NOLOCK) 
    ON o.object_id = c.object_id 
    INNER HASH JOIN sys.columns as c2 with (NOLOCK) 
    ON o.object_id = c2.object_id 
    GROUP BY o.object_id 
) AS x 
    OPTION (HASH GROUP); 

如果我用這個變化，它使用的，而不是MAX TOP，它也適用：

SELECT TOP (1) x.match FROM (
SELECT 
    count(*) as match 
    FROM sys.objects AS o with (nolock, forceseek) 
    INNER LOOP JOIN sys.columns as c with (NOLOCK) 
    ON o.object_id = c.object_id 
    INNER HASH JOIN sys.columns as c2 with (NOLOCK) 
    ON o.object_id = c2.object_id 
    GROUP BY o.object_id 
) AS x 
    ORDER BY x.match DESC 
    OPTION (HASH GROUP); 

也許甚至嘗試消除子查詢（假設僅用於隔離最大值 - 如果沒有必要）：

SELECT TOP (1) 
    count(*) as [max] 
    FROM sys.objects AS o with (nolock, forceseek) 
    INNER LOOP JOIN sys.columns as c with (NOLOCK) 
    ON o.object_id = c.object_id 
    INNER HASH JOIN sys.columns as c2 with (NOLOCK) 
    ON o.object_id = c2.object_id 
    GROUP BY o.object_id 
    ORDER BY [max] DESC 
    OPTION (HASH GROUP); 

我有一個問題，看看這是否是一個類似的問題，我已經報告了對SQL Server 2012.我會更新這裏，當我知道更多。