sql-server - サブクエリのパフォーマンスを最適化する

Question

テーブルからの最初のクエリには約 40 秒かかり、80,000 を超える行が作成されます。サイト、シーケンス、合計、および任意の OS バージョンごとに Windows 7 アプリケーションの数を取得したいと考えています。

これらのサブクエリは機能しますが、もちろんプロセスが大幅に遅くなります。実行に 3.5 時間かかりました。

これを行うより効率的な方法はありますか？

出力：

SoftwareName    Sequence    Site    Win7/site       Win7Installs/seq    TotWin7apps TotalInstalls
Adobe Acrobat       1       BKN         1                   5               626         7854
AutoCAD LT          1       BKN         1                   1               3           15
Adobe Acrobat       1       CTW         4                   5               626         7854
Adobe Captivate     1       CTW         1                   1               8           60

クエリ:

WITH PCapps AS (
SELECT DISTINCT
    Computer,
    Sequence,
    Site,
    SoftwareName,
    OS
FROM table
)

SELECT DISTINCT
    SoftwareName,
    Sequence,
    Site,
    (SELECT COUNT(p1.SoftwareName) FROM PCapps p1 WHERE p1.SoftwareName = pc.Softwarename AND OS = 'Windows 7 Enterprise' AND p1.Site = pc.Site) as 'Win7/site',
    (SELECT COUNT(p1.SoftwareName) FROM PCapps p1 WHERE p1.SoftwareName = pc.Softwarename AND OS = 'Windows 7 Enterprise' AND p1.Sequence = pc.Sequence) as 'Win7Installs/seq',
    (SELECT COUNT(p2.SoftwareName) FROM PCapps p2 WHERE p2.SoftwareName = pc.Softwarename AND OS = 'Windows 7 Enterprise') as TotWin7apps,
    (SELECT COUNT(p3.SoftwareName) FROM PCapps p3 WHERE p3.SoftwareName = pc.SoftwareName) as TotalInstalls 
    FROM PCapps pc

** アップデート：

@Jason Carter のヒントを使用して、いくつかの #temp テーブルを作成し、それらを結合しました。結果は 1 分未満で非常に高速に実行されます (最初のクエリよりも長くはありません)。#temp テーブルを作成するための最初のヒントに従っていたため、この方法は @JasonCarter のソリューションとは少し異なります。それぞれが COUNT() を含むいくつかの #temp テーブルを作成しました。

SELECT DISTINCT
    Computer,
    Sequence,
    Site,
    SoftwareName,
    OS
INTO #PCapps
      FROM TABLE    

SELECT
    SoftwareName,
    Site,
    COUNT(SoftwareName) AS [SiteInstalls]
INTO #SiteInstalls
FROM #PCapps 
    WHERE OS = 'Windows 7 Enterprise'
    GROUP BY Site, SoftwareName

SELECT
    SoftwareName,
    Sequence,
    COUNT(SoftwareName) AS [SeqInstalls]
INTO #SeqInstalls
FROM #PCapps
    WHERE OS = 'Windows 7 Enterprise'
    GROUP BY Sequence, SoftwareName

SELECT
    SoftwareName,
    COUNT(SoftwareName) AS [Win7Installs]
INTO #Win7Installs
FROM #PCapps
    WHERE OS = 'Windows 7 Enterprise'
    GROUP BY SoftwareName

SELECT
    SoftwareName,
    COUNT(SoftwareName) AS [AppInstalls]
INTO #AppInstalls
FROM #PCapps
    GROUP BY SoftwareName   

SELECT
    pc.SoftwareName,
    pc.Sequence,
    pc.Site,
    sit7.SiteInstalls,
    seq7.SeqInstalls,
    w7.Win7Installs,
    ai.AppInstalls
    FROM #PCapps pc
    LEFT OUTER JOIN #SiteInstalls sit7 ON sit7.SoftwareName = pc.SoftwareName AND sit7.Site = pc.Site
    LEFT OUTER JOIN #SeqInstalls  seq7 ON seq7.SoftwareName = pc.SoftwareName AND seq7.Sequence = pc.Sequence
    LEFT OUTER JOIN #Win7Installs w7   ON   w7.SoftwareName = pc.SoftwareName
    LEFT OUTER JOIN #AppInstalls ai    ON   ai.Softwarename = pc.SoftwareName

DROP TABLE #PCapps
DROP TABLE #SiteInstalls
DROP TABLE #SeqInstalls
DROP TABLE #Win7Installs
DROP TABLE #AppInstalls

score 1 · Accepted Answer

最初にサブクエリを一時テーブルに実行してカウントを収集してから、合計カウントを取得してみます。クエリのセットアップ方法では、PCapps のすべての行に対してこれらのサブクエリを 1 回ずつ実行するため、非常に時間がかかります。

次のようなことを試してください：

declare @t table (Computer varchar(10), Sequence int, Site varchar(3), SoftwareName varchar(20), OS varchar(20) )

insert into @t values
 ('C1',1,'BKN','Adobe Acrobat','Win7')
,('C2',1,'BKN','Adobe Acrobat','Win7')
,('C3',1,'BKN','Adobe Acrobat','Win7')
,('C1',1,'BKN','AutoCAD LT ','Win7')
,('C3',1,'BKN','AutoCAD LT ','Win7')
,('B1',2,'CTW','Adobe Acrobat','Win7')
,('B2',2,'CTW','Adobe Acrobat','Win7')
,('B3',3,'CTW','Adobe LT','Win7')
,('B4',3,'CTW','Adobe Acrobat','Win7')
,('A1',2,'CTW','Adobe Acrobat','Win7')
,('A2',2,'CTW','Adobe LT','Win7')
,('A3',2,'CTW','Adobe Acrobat','Win7')
,('X4',3,'CTW','Adobe Acrobat','WinXP')
,('X1',2,'CTW','Adobe Acrobat','WinXP')
,('X2',2,'CTW','Adobe LT','WinXP')
,('X3',2,'CTW','Adobe Acrobat','WinXP')
,('A4',2,'CTW','Adobe Acrobat','Win7');

SELECT 
    Site,
    OS,
    SoftwareName, 
    Sequence,
    COUNT(SoftwareName)  as 'Count'
into #SiteSpecific
FROM @t
group by Site, OS, SoftwareName, Sequence;


WITH PCapps AS (
SELECT DISTINCT
    Computer,
    Sequence,
    Site,
    SoftwareName,
    OS
FROM @t
)
SELECT 
    DISTINCT
    PC.SoftwareName,
    PC.Sequence,
    PC.Site,
    PerSeq.[Count] as 'Win7Installs/seq',
    PerSite.[Count] as TotWin7apps,
    total.[Count] as TotalInstalls 
FROM PCapps pc
inner join (select SoftwareName, Sequence, OS, sum([Count]) [Count] from #SiteSpecific group by SoftwareName, Sequence, OS) PerSeq 
    on PerSeq.SoftwareName = pc.Softwarename 
    AND PerSeq.Sequence = pc.Sequence
    AND PerSeq.OS = pc.OS
inner join (select SoftwareName, Site, OS, sum([Count]) [Count] from #SiteSpecific group by SoftwareName, Site, OS) PerSite 
    on PerSite.SoftwareName = pc.Softwarename 
    AND PerSite.Site = pc.Site
    AND PerSite.OS = pc.OS
inner join (select SoftwareName, sum([Count]) [Count] from #SiteSpecific group by SoftwareName) Total 
    on Total.SoftwareName = pc.Softwarename 
where Pc.OS='Win7'
order by SoftwareName, Sequence, Site

score 0 · Accepted Answer

私には本当に驚くべきことです...オプティマイザは並列処理を利用して、サブクエリを非常に効率的にします。ダミーテーブルに 120K レコードを入力し、以下を試しました。3 番目のクエリは、最初のクエリ (自分のクエリ) よりもわずかに効率的ですが、より複雑になります。私はあなたのものをそのまま残します。おそらくもっと良い解決策がありますが、あなたのものは私には十分に見えます。80K 行を処理するのにどれくらいの時間がかかりますか?

WITH PCapps2 AS (
SELECT DISTINCT
    Computer,
    Sequence,
    Site,
    SoftwareName,
    OS
FROM pcapps
)
SELECT DISTINCT
    SoftwareName,
    Sequence,
    Site,
    (SELECT COUNT(p1.SoftwareName) FROM PCapps2 p1 WHERE p1.SoftwareName = pc.Softwarename AND OS = 'Windows 7 Enterprise' AND p1.Site = pc.Site) as 'Win7/site',
    (SELECT COUNT(p1.SoftwareName) FROM PCapps2 p1 WHERE p1.SoftwareName = pc.Softwarename AND OS = 'Windows 7 Enterprise' AND p1.Sequence = pc.Sequence) as 'Win7Installs/seq',
    (SELECT COUNT(p2.SoftwareName) FROM PCapps2 p2 WHERE p2.SoftwareName = pc.Softwarename AND OS = 'Windows 7 Enterprise') as TotWin7apps,
    (SELECT COUNT(p3.SoftwareName) FROM PCapps2 p3 WHERE p3.SoftwareName = pc.SoftwareName) as TotalInstalls 
    FROM PCapps2 pc

;WITH PCapps2 AS (
SELECT DISTINCT
    Computer,
    Sequence,
    Site,
    SoftwareName,
    OS
FROM pcapps
)
SELECT DISTINCT
    SoftwareName,
    Sequence,
    Site,
    COUNT(case when os = 'Windows 7 Enterprise' then 1 end) over (partition by pc.Softwarename,pc.Site) as 'Win7/site',
    COUNT(case when os = 'Windows 7 Enterprise' then 1 end) over (partition by pc.Softwarename,pc.Sequence) as 'Win7Installs/seq',
    COUNT(case when os = 'Windows 7 Enterprise' then 1 end) over (partition by pc.Softwarename) as TotWin7apps,
    COUNT(*) over (partition by pc.Softwarename) as TotalInstalls
    FROM PCapps2 pc

;WITH PCapps2 AS (
SELECT DISTINCT
    Computer,
    Sequence,
    Site,
    SoftwareName,
    OS
FROM pcapps
), a as (
    select softwarename, site, COUNT(*) as 'Win7/site'
    from pcapps2
    where os = 'Windows 7 Enterprise'
    group by softwarename, site
), b as (
    select softwarename, sequence, COUNT(*) as 'Win7Installs/seq'
    from pcapps2
    where os = 'Windows 7 Enterprise'
    group by softwarename, sequence
), c as (
    select softwarename, COUNT(case when os = 'Windows 7 Enterprise' then 1 end) as TotWin7apps, COUNT(*) as TotalInstalls
    from pcapps2
    group by softwarename
), d as (
    select distinct softwarename, sequence, site
    from    pcapps2
)
select d.*, isnull(a.[Win7/site], 0) as [Win7/site], isnull(b.[Win7Installs/seq],0) as [Win7Installs/seq] 
        , isnull(c.TotWin7apps,0) as TotWin7apps, isnull(c.TotalInstalls, 0) as TotalInstalls  
from    d
        left join a on d.softwarename = a.softwarename and d.site = a.site 
        left join b on d.softwarename = b.softwarename and d.sequence = b.sequence 
        left join c on d.softwarename = c.softwarename

最初のクエリ: テーブル 'pcapps'。スキャンカウント 15、論理読み取り 6630、物理読み取り 0、先読み読み取り 0、LOB 論理読み取り 0、LOB 物理読み取り 0、LOB 先読み読み取り 0。スキャンカウント 0、論理読み取り 0、物理読み取り 0、先読み読み取り 0、LOB 論理読み取り 0、LOB 物理読み取り 0、LOB 先読み読み取り 0。

2 番目のクエリ: テーブル 'pcapps'。スキャンカウント 3、論理読み取り 1326、物理読み取り 0、先読み読み取り 0、LOB 論理読み取り 0、LOB 物理読み取り 0、LOB 先読み読み取り 0。スキャンカウント 18、論理読み取り 1983591、物理読み取り 0、先読み読み取り 0、LOB 論理読み取り 0、LOB 物理読み取り 0、LOB 先読み読み取り 0。

3 番目のクエリ: テーブル 'pcapps'。スキャンカウント 12、論理読み取り 5304、物理読み取り 0、先読み読み取り 0、LOB 論理読み取り 0、LOB 物理読み取り 0、LOB 先読み読み取り 0。スキャンカウント 0、論理読み取り 0、物理読み取り 0、先読み読み取り 0、LOB 論理読み取り 0、LOB 物理読み取り 0、LOB 先読み読み取り 0。

sql-server - サブクエリのパフォーマンスを最適化する

2 に答える 2

Related

Reference