選擇執行時間太長的查詢（SQL Server 2008）

Question

上週我試圖運行這個查詢，但在7小時後它仍在執行，我不得不在一天結束時取消。它只返回了800多行，最終結果應該是20000行。任何提高性能的建議都非常值得讚賞。

非常感謝。

Select 
    a.entryID, g.GroupName, 
    MAX(CASE WHEN e.attributedefinitionid = 15 THEN e.AttributeValue END) AS EnglishWording, 
    MAX(CASE WHEN e.attributedefinitionid = 41 THEN e.AttributeValue END) AS GermanWording, 
    MAX(CASE WHEN e.attributedefinitionid = 64 THEN e.AttributeValue END) AS RussianWording, 
    MAX(CASE WHEN e.attributedefinitionid = 91 THEN e.AttributeValue END) AS FrenchWording, 
    MAX(CASE WHEN e.attributedefinitionid = 143 THEN e.AttributeValue END) AS ItalianWording, 
    MAX(CASE WHEN e.attributedefinitionid = 160 THEN e.AttributeValue END) AS SpanishWording, 
    MAX(CASE WHEN e.attributedefinitionid = 165 THEN e.AttributeValue END) AS ChineseWording, 
    MAX(CASE WHEN e.attributedefinitionid = 170 THEN e.AttributeValue END) AS JapaneseWording, 
    MAX(CASE WHEN h.attributedefinitionid = 17 
       THEN convert(varchar(10), h.AttributeValue, 120) 
     END) as DateA, 
    MAX(CASE WHEN h.attributedefinitionid = 557 
       THEN convert(varchar(10), h.AttributeValue, 120) 
     END) as DateO, 
    MAX(CASE WHEN h.attributedefinitionid = 558 
       THEN convert(varchar(10), h.AttributeValue, 120) 
     END) as DateC, 
    MAX(CASE WHEN j.attributedefinitionid = 29 THEN j.AttributeValue END) AS Warning, 
    MAX(CASE WHEN l.attributedefinitionid = 23 THEN l.AttributeValue END) AS Highlight, 
    MAX(CASE WHEN n.attributedefinitionid = 572 THEN n.AttributeValue END) AS Investigations, 
    MAX(CASE WHEN p.attributedefinitionid = 30 THEN p.AttributeValue END) AS Comments, 
    MAX(CASE WHEN s.attributedefinitionid = 39 THEN s.AttributeValue END) AS USD, 
    MAX(CASE WHEN s.attributedefinitionid = 40 THEN s.AttributeValue END) AS EUR, 
    MAX(CASE WHEN s.attributedefinitionid = 92 THEN s.AttributeValue END) AS GBP, 
    MAX(CASE WHEN s.attributedefinitionid = 450 THEN s.AttributeValue END) AS RBN, 
    MAX(CASE WHEN s.attributedefinitionid = 451 THEN s.AttributeValue END) AS JPY, 
    MAX(CASE WHEN s.attributedefinitionid = 552 THEN s.AttributeValue END) AS HK$, 
    MAX(CASE WHEN u.attributeDefinitionID = 142 THEN u.attributeValue END) AS [Status] 
From 
    Entry as a 
inner join 
    entrycategory as b on b.entryid = a.entryid 
inner join 
    category as c on c.categoryid = b.categoryid 
inner join 
    [Group] as g on g.groupID = c.groupID 
left outer join 
    EntryAttribute AS d on d.entryID = A.entryid 
left outer join 
    attributestring AS e on e.attributeid = d.attributeid 
left outer join 
    EntryAttribute as f on f.EntryID = a.EntryID 
left outer join 
    AttributeDateTime as h on h.AttributeID = f.AttributeID 
left outer join 
    EntryAttribute as i on i.EntryID = a.EntryID 
left outer join 
    AttributeString as j on j.AttributeID = i.AttributeID 
left outer join 
    EntryAttribute as k on k.EntryID = a.EntryID 
left outer join 
    AttributeString as l on l.AttributeID = k.AttributeID 
left outer join 
    EntryAttribute as m on m.EntryID = a.EntryID 
left outer join 
    AttributeString as n on n.AttributeID = m.AttributeID 
left outer join 
    EntryAttribute as o on o.EntryID = a.EntryID 
left outer join 
    AttributeString as p on p.AttributeID = o.AttributeID 
inner join 
    EntryAttribute AS r on r.entryID = A.entryid 
inner join 
    attributestring AS s on s.attributeid = r.attributeid 
left outer join 
    EntryAttribute as t on t.EntryID = a.EntryID 
left outer join 
    AttributeString as u on u.AttributeID = t.AttributeID 
Group by 
    a.entryID, g.GroupName 

Answer 1

據我所知，你只需要一個加入到EntryAttribute和AttributeString表。聚合照顧其餘部分。

Select a.entryID,g.GroupName, 
MAX(CASE WHEN e.attributedefinitionid = 15 THEN e.AttributeValue END) AS EnglishWording, 
MAX(CASE WHEN e.attributedefinitionid = 41 THEN e.AttributeValue END) AS GermanWording, 
MAX(CASE WHEN e.attributedefinitionid = 64 THEN e.AttributeValue END) AS RussianWording, 
MAX(CASE WHEN e.attributedefinitionid = 91 THEN e.AttributeValue END) AS FrenchWording, 
MAX(CASE WHEN e.attributedefinitionid = 143 THEN e.AttributeValue END) AS ItalianWording, 
MAX(CASE WHEN e.attributedefinitionid = 160 THEN e.AttributeValue END) AS SpanishWording, 
MAX(CASE WHEN e.attributedefinitionid = 165 THEN e.AttributeValue END) AS ChineseWording, 
MAX(CASE WHEN e.attributedefinitionid = 170 THEN e.AttributeValue END) AS JapaneseWording, 

MAX(CASE WHEN e.attributedefinitionid = 17 THEN convert(varchar(10),e.AttributeValue,120) END) as DateA, 
MAX(CASE WHEN e.attributedefinitionid = 557 THEN convert(varchar(10),e.AttributeValue,120) END) as DateO, 
MAX(CASE WHEN e.attributedefinitionid = 558 THEN convert(varchar(10),e.AttributeValue,120) END) as DateC, 

MAX(CASE WHEN e.attributedefinitionid = 29 THEN e.AttributeValue END) AS Warning, 
MAX(CASE WHEN e.attributedefinitionid = 23 THEN e.AttributeValue END) AS Highlight, 
MAX(CASE WHEN e.attributedefinitionid = 572 THEN e.AttributeValue END) AS Investigations, 
MAX(CASE WHEN e.attributedefinitionid = 30 THEN e.AttributeValue END) AS Comments, 

MAX(CASE WHEN e.attributedefinitionid = 39 THEN e.AttributeValue END) AS USD, 
MAX(CASE WHEN e.attributedefinitionid = 40 THEN e.AttributeValue END) AS EUR, 
MAX(CASE WHEN e.attributedefinitionid = 92 THEN e.AttributeValue END) AS GBP, 
MAX(CASE WHEN e.attributedefinitionid = 450 THEN e.AttributeValue END) AS RBN, 
MAX(CASE WHEN e.attributedefinitionid = 451 THEN e.AttributeValue END) AS JPY, 
MAX(CASE WHEN e.attributedefinitionid = 552 THEN e.AttributeValue END) AS HK$, 
MAX(CASE WHEN e.attributeDefinitionID = 142 THEN e.attributeValue END) AS [Status] 

From Entry as a 
inner join entrycategory as b on b.entryid = a.entryid 
inner join category as c on c.categoryid= b.categoryid 
inner join [Group] as g on g.groupID = c.groupID 
left outer join EntryAttribute AS d on d.entryID = A.entryid 
left outer join attributestring AS e on e.attributeid = d.attributeid 
Group by a.entryID,g.GroupName; 

通過具有多個連接，則在每次entryId，GroupName組內產生humungous笛卡爾乘積。