MS SQL 日常維護(hù)管理常用腳本(二)

鷹皇軟件 2013-08-14

展開全文

監(jiān)控數(shù)據(jù)庫運(yùn)行

下面是整理、收集監(jiān)控數(shù)據(jù)庫運(yùn)行的一些常用腳本，也是MS SQL 日常維護(hù)管理常用腳本(一)的續(xù)集，歡迎大家補(bǔ)充、提意見。

查看數(shù)據(jù)庫登錄名信息

Code Snippet

SELECT name                                AS LoginName ,
       dbname                              AS DefaultDB ,
       createdate                          AS CreateDate,
       updatedate                          AS UpdateDate,
       language                            AS Language  ,
       CASE WHEN isntname = 1 THEN 'NT USER'
           ELSE 'SQL USER'    END          AS UserType
FROM syslogins;

查看數(shù)據(jù)庫用戶信息

SELECT * FROM sysusers;

查看用戶擁有的服務(wù)器角色

方法1：用SSMS管理工具查看

方法2：腳本查詢

查看用戶角色

SELECT name            ,
       CASE WHEN sysadmin     = 1       THEN 'yes'     ELSE '' END AS IsSysadmin        ,
       CASE WHEN dbcreator    = 1       THEN 'yes'     ELSE '' END AS IsDbCreate        ,
       CASE WHEN securityadmin= 1       THEN 'yes'     ELSE '' END AS IsSecurityadmin   ,
       CASE WHEN bulkadmin    = 1       THEN 'yes'     ELSE '' END AS IsBulkadmin       ,
       CASE WHEN diskadmin    = 1       THEN 'yes'     ELSE '' END AS IsDiskadmin       ,
       CASE WHEN processadmin = 1       THEN 'yes'     ELSE '' END AS IsProcessadmin    ,
       CASE WHEN serveradmin  = 1       THEN 'yes'     ELSE '' END AS IsServeradmin     ,
       CASE WHEN setupadmin   = 1       THEN 'yes'     ELSE '' END AS IsSetupadmin    
FROM syslogins
--WHERE NAME='loginname'

查看最大工作線程數(shù)

Code Snippet

SELECT  max_workers_count
  FROM  sys.dm_os_sys_info

查看當(dāng)前用戶進(jìn)程的會話ID

SELECT @@SPID

查詢當(dāng)前會話使用哪種協(xié)議

Code Snippet

SELECT net_transport
FROM   sys.dm_exec_connections
WHERE session_id = @@SPID;

查看當(dāng)前連接的會話信息

--進(jìn)程號1--50是SQL Server系統(tǒng)內(nèi)部用的

SELECT * FROM sys.dm_exec_sessions WHERE session_id >=51

--查看某臺機(jī)器的連接會話信息

SELECT * FROM sys.dm_exec_sessions WHERE session_id >=51 AND host_name='PO130018801'

--查看某個登錄名的連接會話信息

SELECT * FROM sys.dm_exec_sessions WHERE session_id >=51 AND login_name='username'

--查看活動的連接會話信息

SELECT * FROM sys.dm_exec_sessions WITH(NOWAIT) WHERE session_id >=51 AND status ='running'

--查找連接到服務(wù)器的用戶并返回每個用戶的會話數(shù)

SELECT  login_name ,

        COUNT(session_id) AS session_count

FROM    sys.dm_exec_sessions

GROUP BY login_name ;

查看正在執(zhí)行的SQL語句

方法1：選擇數(shù)據(jù)庫實(shí)例，單擊右鍵，選擇”活動監(jiān)視器“，監(jiān)控/查看正在執(zhí)行的SQL

方法2： profile去跟蹤，比較耗費(fèi)資源。

方法3：

Code Snippet

SELECT[Spid] = session_Id ,
      ecid ,
      [Database] = DB_NAME(sp.dbid) ,
      [User] = nt_username ,
      [Status] = er.status ,
      [Wait] = wait_type ,
      [Individual Query] = SUBSTRING(qt.text, er.statement_start_offset / 2,
                                     ( CASE WHEN er.statement_end_offset = -1
                                            THEN LEN(CONVERT(NVARCHAR(MAX), qt.text))
                                                 * 2
                                            ELSE er.statement_end_offset
                                       END - er.statement_start_offset ) / 2) ,
      [Parent Query] = qt.text ,
      Program = program_name ,
      Hostname ,
      nt_domain ,
      start_time
  FROMsys.dm_exec_requests er
      INNER JOIN sys.sysprocesses sp ON er.session_id = sp.spid
      CROSS APPLY sys.dm_exec_sql_text(er.sql_handle) AS qt
 WHEREsession_Id >= 51

方法4：

Code Snippet

SELECT   m.session_id ,
         m.start_time ,
         m.command    ,
         m.wait_type  ,
         m.cpu_time   ,
        CAST(s.text AS VARCHAR(1000)) AS sqlText
 FROMmaster.sys.dm_exec_requests m WITH ( NOLOCK )
     CROSS APPLY fn_get_sql(m.sql_handle) s
 
SELECT  r.session_id,
        r.start_time      ,
        r.command         ,
        r.wait_type       ,
        r.cpu_time        ,
        s.text  
 FROMsys.dm_exec_requests r
     CROSS APPLY sys.dm_exec_sql_text(r.sql_handle) s

--查看某個會話ID正在執(zhí)行的SQL

Code Snippet

SELECT  m.session_id ,
        m.start_time ,
        m.command    ,
        m.wait_type  ,
        m.cpu_time   ,
      CAST(s.text AS VARCHAR(1000)) AS sqlText
  FROMmaster.sys.dm_exec_requests m WITH ( NOLOCK )
      CROSS APPLY fn_get_sql(m.sql_handle) s
 WHEREm.session_id = 342
 
    SELECT   r.session_id      ,
             r.start_time      ,
             r.command         ,
             r.wait_type       ,
             r.cpu_time        ,
             s.text  
  FROM sys.dm_exec_requests r
      CROSS APPLY sys.dm_exec_sql_text(r.sql_handle) s
WHERE r.seesion_id =342

查看SQL SERVER進(jìn)程執(zhí)行的語句

Code Snippet

USE master
 
DECLARE @spid INT ;
 
DECLARE @sql_handle BINARY(20) ;
 
SET @spid = 56
 
  SELECT@sql_handle = sql_handle
    FROMsysprocesses AS A WITH ( NOLOCK )
   WHEREspid = @spid ;
 
  SELECTtext
    FROM::fn_get_sql(@sql_handle) ;

查找TOP N語句

按平均 CPU 時間返回排名前十個的查詢的相關(guān)信息。此示例將根據(jù)查詢的查詢哈希對查詢進(jìn)行聚合，以便按照查詢的累積資源消耗來分組在邏輯上等效的查詢。
--注意：SQL 2005 某些版本，沒有sys.dm_exec_query_stats系統(tǒng)動態(tài)視圖沒有query_hash視圖。

Code Snippet

USE DBNAME;
GO
SELECT TOP 10 query_stats.query_hash             AS "Query Hash",
    SUM(query_stats.total_worker_time) /
    SUM(query_stats.execution_count)             AS "Avg CPU Time",
    MIN(query_stats.statement_text)              AS "Statement Text"
FROM
    (SELECT QS.*,
    SUBSTRING(ST.text,(QS.statement_start_offset/2) + 1,
    ((CASE statement_end_offset
        WHEN -1 THEN DATALENGTH(st.text)
        ELSE QS.statement_end_offset END
            - QS.statement_start_offset)/2) + 1) AS statement_text
     FROM sys.dm_exec_query_stats AS QS
     CROSS APPLY sys.dm_exec_sql_text(QS.sql_handle) as ST) as query_stats
GROUP BY query_stats.query_hash
ORDER BY 2 DESC;
GO

查看會話阻塞/死鎖信息

方法1：查看那個引起阻塞，查看blk不為0的記錄，如果存在阻塞進(jìn)程，則是該阻塞進(jìn)程的會話 ID。否則該列為零。

EXEC sp_who active

方法2：查看那個引起阻塞，查看字段BlkBy，這個能夠得到比sp_who更多的信息。

EXEC sp_who2 active

方法3：sp_lock 系統(tǒng)存儲過程，報告有關(guān)鎖的信息，但是不方便定位問題

方法4：sp_who_lock存儲過程

Code Snippet

USE master;
GO
 
SET ANSI_NULLS ON;
GO
 
SET QUOTED_IDENTIFIER ON;
GO
 
IF EXISTS(SELECT 1 FROM sysobjects WHERE id=OBJECT_ID(N'sp_who_lock') 
                                       AND OBJECTPROPERTY(id, 'IsProcedure') =1)
    DROP PROCEDURE sp_who_lock;
GO
 
--==================================================================================================
--            ProcedureName             :            sp_who_lock
--            Author                    :            作者不詳,出自網(wǎng)絡(luò)
--            CreateDate                :            2013-05-13
--            Description               :            查看阻塞和死鎖信息
/**************************************************************************************************
        Parameters                    :              參數(shù)說明
***************************************************************************************************
                                                     無參存儲過程
***************************************************************************************************
        Modified Date            Modified User     Version                 Modified Reason
***************************************************************************************************
       2013-06-03                    Kerry        V01.00.01 調(diào)整存儲過程格式,代碼部分修改以及增加注釋信息
***************************************************************************************************/
--=================================================================================================
 
  CREATE PROCEDURE sp_who_lock
    AS
BEGIN
 
DECLARE @spid                            INT;
DECLARE @block                           INT;
DECLARE @RowCount                        INT;
DECLARE @RowIndex                        INT;
 
--創(chuàng)建臨時表，保持被阻塞或正阻塞其他SQL的SQL語句信息
CREATE TABLE #tmp_lock_who
(
      id    INT IDENTITY(1, 1) ,
      spid    SMALLINT ,
      block SMALLINT
)
    
    
IF @@ERROR<>0 RETURN @@ERROR;
 
INSERT INTO #tmp_lock_who
        (
          spid ,
          block
        )
        SELECT  0 ,
                blocked
        FROM    ( SELECT    *
                  FROM      sysprocesses
                  WHERE     blocked > 0
                ) a
        WHERE   NOT EXISTS( SELECT *
                             FROM   ( SELECT    *
                                      FROM      sysprocesses
                                      WHERE     blocked > 0
                                    ) b
                             WHERE  a.blocked = spid )
        UNION
        SELECT  spid ,
                blocked
        FROM    sysprocesses
        WHERE   blocked > 0;
        
        
IF @@ERROR<>0 RETURN @@ERROR;
 
 
-- 找到臨時表的記錄數(shù)
  SELECT@RowCount = COUNT(1) ,
        @RowIndex = 1
    FROM#tmp_lock_who
 
 
IF @@ERROR<>0 RETURN @@ERROR;
 
  IF@RowCount=0
  SELECT  N'現(xiàn)在沒有阻塞和死鎖信息' AS MESSAGE;
 
 
-- -- 循環(huán)開始
WHILE @RowIndex <= @RowCount
BEGIN
    -- 取第一條記錄
    SELECT     @spid    = spid,
               @block   = block
    FROM #tmp_lock_who
    WHERE Id = @RowIndex
 
    IF @spid = 0
        SELECT N'引起數(shù)據(jù)庫死鎖的是: ' + CAST(@block AS VARCHAR(10))
            +  N'進(jìn)程號,其執(zhí)行的SQL語法如下';
    ELSE
        SELECT N'進(jìn)程號SPID：' + CAST(@spid AS VARCHAR(10)) + N'被進(jìn)程號SPID：'
            +  CAST(@block AS VARCHAR(10)) +'阻塞,其當(dāng)前進(jìn)程執(zhí)行的SQL語法如下';
        
    DBCC INPUTBUFFER(@block )
 
 
    SET @RowIndex = @RowIndex + 1;
 
END;
 
    DROP TABLE #tmp_lock_who;
 
    RETURN 0;
END

方法5：右鍵服務(wù)器-選擇“活動和監(jiān)視器”，查看進(jìn)程選項(xiàng)。注意“任務(wù)狀態(tài)”字段。

方法6：右鍵服務(wù)名稱-選擇報表-標(biāo)準(zhǔn)報表-活動-所有正在阻塞的事務(wù)。

小結(jié)：總結(jié)之后，才發(fā)現(xiàn)居然有這么多方法，MGD，系統(tǒng)的整理、梳理知識點(diǎn)是非常有必要的，你能更全面、深入的了解。

查看內(nèi)存狀態(tài)

dbcc memorystatus

具體如何分析，請查看官方文檔http://support.microsoft.com/kb/907877/zh-cn

查看腳本執(zhí)行時間

方法1：查看SSMS管理器，查詢窗口右下角
方法2:

Code Snippet

DECLARE @exectime DATETIME
  SELECT@exectime = GETDATE()
--SQL 語句
PRINT N'SQL執(zhí)行耗時:' + CONVERT(VARCHAR(10), DATEDIFF(ms, @exectime, GETDATE()))

方法3：

SET STATISTICS TIME ON

--SQL 語句

查看進(jìn)程正在執(zhí)行的SQL語句

　　dbcc inputbuffer ()

查看那些表缺少索引

下面語句功能強(qiáng)大，執(zhí)行結(jié)果受統(tǒng)計信息的影響

Code Snippet

SELECT sys.objects.name table_name, 
       mid.statement full_name, 
    (migs.avg_total_user_cost * migs.avg_user_impact) *(migs.user_seeks + migs.user_scans) AS Impact, 
    migs.avg_user_impact *(migs.user_seeks + migs.user_scans) Avg_Estimated_Impact,
    'CREATE NONCLUSTERED INDEX IDX_' + sys.objects.name + '_N ON ' 
        + sys.objects.name COLLATE DATABASE_DEFAULT 
        + ' ( ' + IsNull(mid.equality_columns, '') 
        + CASE WHEN mid.inequality_columns IS NULL 
                THEN '' 
            ELSE 
                CASE WHEN mid.equality_columns IS NULL
                    THEN '' 
                ELSE ',' 
                END + mid.inequality_columns 
            END + ' ) ' 
        + CASE WHEN mid.included_columns IS NULL
                THEN '' 
            ELSE 'INCLUDE (' + mid.included_columns + ')' END 
        + ';' AS CreateIndexStatement, 
    mid.equality_columns, 
    mid.inequality_columns, 
    mid.included_columns 
FROM sys.dm_db_missing_index_group_stats AS migs 
    INNER JOIN sys.dm_db_missing_index_groups AS mig ON migs.group_handle = mig.index_group_handle
    INNER JOIN sys.dm_db_missing_index_details AS mid ON mig.index_handle = mid.index_handle 
        AND mid.database_id = DB_ID()
    INNER JOIN sys.objects WITH (nolock) ON mid.OBJECT_ID = sys.objects.OBJECT_ID
WHERE (migs.group_handle IN
        (
            SELECT TOP (500) group_handle
            FROM sys.dm_db_missing_index_group_stats WITH (nolock)
            ORDER BY (avg_total_user_cost * avg_user_impact) *(user_seeks + user_scans) DESC)) 
    AND OBJECTPROPERTY(sys.objects.OBJECT_ID, 'isusertable')=1
--ORDER BY [Impact] DESC, [full_name] DESC
ORDER BY [table_name], [Impact] desc

查看應(yīng)該被移除的索引

查看那些多余的、應(yīng)該被移除的索引

SQL 1：

Code Snippet

SELECT OBJECTNAME = OBJECT_NAME(I.OBJECT_ID),
    INDEXNAME = I.NAME,
    I.INDEX_ID
FROM SYS.INDEXES I
    JOIN SYS.OBJECTS O ON I.OBJECT_ID = O.OBJECT_ID
WHERE OBJECTPROPERTY(O.OBJECT_ID,'IsUserTable') = 1
    AND I.INDEX_ID NOT IN(
                            SELECT S.INDEX_ID
                            FROM SYS.DM_DB_INDEX_USAGE_STATS S
                            WHERE S.OBJECT_ID = I.OBJECT_ID
                                AND I.INDEX_ID = S.INDEX_ID
                                AND DATABASE_ID = DB_ID())
ORDER BY OBJECTNAME, I.INDEX_ID, INDEXNAME ASC

SQL 2: 分析下面語句，移除那些沒有必要的索引

Code Snippet

SELECT DB_NAME(database_id)         AS  N'DataBaseName'  ,
       OBJECT_NAME(U.object_id)     AS  N'Table_Name'    ,
       I.name                       AS  N'Index_Name'    ,
       user_seeks                   AS  N'用戶索引查找次數(shù)',
       user_scans                   AS  N'用戶索引掃描次數(shù)',
       last_user_seek               AS  N'最后查找時間'   ,
       last_user_scan               AS  N'最后掃描時間'   ,
       rows                         AS  N'表中的行數(shù)'
FROM sys.dm_db_index_usage_stats AS U
   INNER JOIN sys.indexes I ON U.index_id= I.index_idAND U.object_id= I.object_id
  INNER JOIN sysindexesT ON I.object_id = T.id
WHERE database_id= DB_ID('DbName')
AND OBJECT_NAME(U.object_id)='TableName'
ORDER BY user_seeks, user_scans, object_name(U.object_id);

作者：瀟湘隱者

本站是提供個人知識管理的網(wǎng)絡(luò)存儲空間，所有內(nèi)容均由用戶發(fā)布，不代表本站觀點(diǎn)。請注意甄別內(nèi)容中的聯(lián)系方式、誘導(dǎo)購買等信息，謹(jǐn)防詐騙。如發(fā)現(xiàn)有害或侵權(quán)內(nèi)容，請點(diǎn)擊一鍵舉報。

轉(zhuǎn)藏 分享

QQ空間 QQ好友新浪微博微信

獻(xiàn)花（0） +1

來自：鷹皇軟件 > 《待分類1》

舉報/認(rèn)領(lǐng)