Tag Archives: tutorial

How to size transaction log files appropriately and reduce the number of virtual log files

Sizing transaction log files and resolving the issue of having too many virtual log files may or may not be caused by the same problem but both have similar solutions revolving around the same steps. As such the script provide below can resolve both but the application of the script differs slightly.

Appropriately sizing the Transaction Log (TLog) File

Ideally, TLog files should be created at their desired size at installation rather than being allowed to grow slowly over time.

Typically though the DBA will not know how large or frequent the transactions will be prior to the database becoming active in the production environment.

Or the DBA might be new to the organisation and may not be able to tell if the TLog file size is appropriate for the typical transactions the database handles or is the result of some abnormally large transactions or failed TLog backups.

All this can of course be investigated somewhat but if the database transactions are in the kilobytes and transaction frequency is low then you could also take the pragmatic approach outlined below:

  1. Backup the log until it gets truncated
  2. Shrink the log as much as possible
  3. Allow the file to grow
  4. Record the maximum size of the log
  5. Backup the log until it gets truncated
  6. Shrink the log as much as possible
  7. Manually expand the log to the maximum size recorded in step 4

Carrying out the steps above will also resolve any issues with excessive Virtual Log Files. This is dealt with in more detail below.

The script provided at the end of this article can be utilized to achieve the steps above, see the instructions “How to use the script” below.

Before proceeding with this approach however ensure the following points have been considered as these will minimize the impact of growing the files on the databases:

  • That the autogrowth increment for the transaction log is suitably large, @DefaultFileGrowth is set at 128 Mb. This value or higher is recommended.
  • Appropriate restriction settings for growth, preferably unrestricted
  • Instant file initialization is on.

Review the following article to set up instant file initialization.

https://www.brentozar.com/blitz/instant-file-initialization/

You may already know that the TLog file has been appropriately sized but there are too many VLFs in which case the following section outlines how to resolve VLFs as an issue.

Dealing with Virtual Log Files (VLFs)

Each Transaction Log (TLog) file is divided logically into smaller segments called VLFs. VLFs are the unit of truncation for the TLog. When a VLF no longer contains log records for active transactions, it can be truncated and the space becomes available to log new transactions.

SQL Server should not have an excessive number of VLFs inside the TLog. Having a large number of small VLFs can cause the following problems:

  • Slow down the recovery process on startup or after restoring a backup
  • Slow down TLog backups
  • Affect insert/update/delete performance

What constitutes too many VLFs as a problem is subjective however Microsoft flags 1000+ VLFs as an issue.

The number of VLFs can grow based on the autogrowth settings for the TLog file and how often active transactions are written to disk. The problem of too many VLFs is often the result of accepting the default settings for initial log file size and autogrowth increment when the database is created.

To check how many VLFs a database has run the below command against the database:

DBCC LOGINFO;

Each row returned represents a VLF.

To resolving too many VLFs unfortunately requires some downtime. The fix is to shrink and regrow the log file. This is a blocking operation and I would suggest once complete you run a full backup on the database\s as the shrink will break the continuity of the Full and TLog backup schedule which will cause the TLog backups to fail.

Taking the following steps will reduce the number of VLFs in the TLog of a database:

  1. Record the maximum size of the log
  2. Backup the log until it gets truncated
  3. Shrink the log as much as possible
  4. Manually expand the log to the maximum size recorded in the first step

The script provided at the end of this article can be utilized to achieve the steps above, see the instructions “How to use the script” below.

How to use the script:

The script below will output the TSQL to run against all the databases in your instance. If you want to address every database on the instance run the entire script during scheduled downtime. To run against a specific database simply run the snippet that relates to that database.

The role of Backups

The TLog needs to be backed up and truncated prior to the running the shrink to get the file as small as possible.

Backup Option 1:

The script will backup the TLog\s if you provide a value for the @BackupLocation variable. This is defaulted to ‘C:\SQL_Server\Backup\’.

Backup Option 2: 

Alternatively you can set  @BackupLocation = NULL and run your own TLog backup maintenance plan or maintenance script prior to running the script.

Future Proofing by adding a margin for growth:

The script variable @Margin will allow you to set a percentage of how much larger you want the new TLog file size base on the current size. So for a database with a current TLog file size of 100 Mb, running the script with @Margin = 10 will output TSQL to change the TLog file size to 110 Mb. This has increased the TLog File size by 10%, providing 10% additional disk space before the TLog would need to grow again.

Appropriate TLog File Size Instructions:

Once you have backups arranged using one the options above set the variable @LogSizeMb = ‘128’. And run the script. This will start the TLog file size at 128 Mb.

Take the TSQL Generated and run the snippets needed for particular databases or run the entire script.

Let the file grow over a period that is representative of repeated database transcation activity.

When the TLog file is no longer growing you have found the natural size of the TLog for that database.

Run the script again this time setting @LogSizeMb = 0 and if appropriate provide a value for @Margin. The TSQL generated will then use the current TLog file size as is, plus the optional margin.

Run the snippets for the required database\s.

Dealing with Virtual Log Files (VLFs) Instructions:

Once you have backups arranged using one the options above, ensure that the variable @LogSizeMb is set to zero and if appropriate provide a value for @Margin. Then the script can be run to generate the TSQL needed to shrink and resize the TLog file size reducing the number of VLFs.

Run the snippets for the required database\s.

USE master;
GO

SET QUOTED_IDENTIFIER ON
GO

SET ARITHABORT ON
GO

DECLARE @DbName NVARCHAR(255)
	,@LogicalFileName NVARCHAR(255)
	,@DBRecoveryDesc VARCHAR(200)
	,@LogSizeMb INT
	,@DefaultFileGrowth VARCHAR(10)
	,@BackupLocation VARCHAR(255)
	,@GetDate VARCHAR(23)
	,@Margin AS DECIMAL(5, 2)

SELECT @Margin = 0
	,@LogSizeMb = 0
	,@BackupLocation = 'C:\SQL_Server\Backup\'
	,@DefaultFileGrowth = '128'
	,@GetDate = CONVERT(VARCHAR(23), GETDATE(), 126)
	,@GetDate = REPLACE(REPLACE(REPLACE(REPLACE(@GetDate, '-', ''), ':', ''), 'T', '_'), '.', '')

DECLARE DatabaseList CURSOR
FOR
SELECT d.NAME
	,d.recovery_model_desc AS DBRecoveryDesc
	,CASE 
		WHEN @LogSizeMb > 0
			THEN @LogSizeMb
		ELSE CEILING((mf.size * 8.0) / 1024) + (CEILING(((mf.size * 8.0) / 1024) * (@Margin / 100)))
		END AS LogSizeMb
	,@BackupLocation AS BackupLocation
	,@DefaultFileGrowth AS DefaultFileGrowth
FROM sys.databases AS d
LEFT JOIN sys.master_files AS mf ON d.database_Id = mf.database_Id
WHERE d.state_desc = 'ONLINE'
	AND d.is_read_only = 0
	AND d.database_id > 4
	AND type_desc = 'LOG'
ORDER BY NAME

OPEN DatabaseList

FETCH NEXT
FROM DatabaseList
INTO @DbName
	,@DBRecoveryDesc
	,@LogSizeMb
	,@BackupLocation
	,@DefaultFileGrowth

WHILE @@FETCH_STATUS = 0
BEGIN
	SET @LogicalFileName = (
			SELECT TOP 1 NAME
			FROM sys.master_files AS mf
			WHERE DB_NAME(database_id) = @DbName
				AND type_desc = 'LOG'
			)

	IF @DBRecoveryDesc = 'Full'
		AND @BackupLocation IS NOT NULL
	BEGIN
		PRINT (
				'Use [' + @DbName + '] 
            GO  
			
			BACKUP LOG [' + @DbName + '] TO DISK=''' + @BackupLocation + @DbName + '_' + @GetDate + '_Log.trn' + ''';
			GO
			
            ALTER DATABASE [' + @DbName + '] SET RECOVERY SIMPLE WITH NO_WAIT
            GO   

            DBCC SHRINKFILE (''' + @LogicalFileName + ''',10)  
            GO  
            
            ALTER DATABASE [' + @DbName + '] MODIFY FILE (NAME=''' + @LogicalFileName + ''', SIZE=' + CONVERT(VARCHAR(10), @LogSizeMb) + 'MB, MAXSIZE=UNLIMITED, FILEGROWTH=' + @DefaultFileGrowth + 'MB);
            GO

            ALTER DATABASE [' + @DbName + '] SET RECOVERY FULL WITH NO_WAIT
            GO '
				)
		PRINT '----------------------------------------------------------- '
	END
	ELSE IF @DBRecoveryDesc = 'Full'
	BEGIN
		PRINT (
				'Use [' + @DbName + '] 
            GO  
					
            ALTER DATABASE [' + @DbName + '] SET RECOVERY SIMPLE WITH NO_WAIT
            GO   

            DBCC SHRINKFILE (''' + @LogicalFileName + ''',10)  
            GO  
            
            ALTER DATABASE [' + @DbName + '] MODIFY FILE (NAME=''' + @LogicalFileName + ''', SIZE=' + CONVERT(VARCHAR(10), @LogSizeMb) + 'MB, MAXSIZE=UNLIMITED, FILEGROWTH=' + @DefaultFileGrowth + 'MB);
            GO

            ALTER DATABASE [' + @DbName + '] SET RECOVERY FULL WITH NO_WAIT
            GO '
				)
		PRINT '----------------------------------------------------------- '
	END
	ELSE IF @DBRecoveryDesc = 'Simple'
	BEGIN
		PRINT (
				'Use [' + @DbName + ']  
            GO  

            DBCC SHRINKFILE (''' + @LogicalFileName + ''',10)    
            GO
            
            ALTER DATABASE [' + @DbName + '] MODIFY FILE (NAME=''' + @LogicalFileName + ''', SIZE=' + CONVERT(VARCHAR(10), @LogSizeMb) + 'MB, MAXSIZE=UNLIMITED, FILEGROWTH=' + @DefaultFileGrowth + 'MB);
            GO
            '
				)
		PRINT '----------------------------------------------------------- '
	END

	FETCH NEXT
	FROM DatabaseList
	INTO @DbName
		,@DBRecoveryDesc
		,@LogSizeMb
		,@BackupLocation
		,@DefaultFileGrowth
END

CLOSE DatabaseList

DEALLOCATE DatabaseList

How to capture command prompt output and write it to text file

Say you have a batch script that you run that throws back information you want to keep. Well maybe the best way to do that is to have the output written to a text file. This technique can then be combined with a task scheduler so you have an  automated task that creates a log file.

The below batch script below has two main parts. The first part creates a file to capture the command output. The second part is the command prompt command to run.

Creating the file takes the following 2 parameters:

  • filePath: Where you want the file written, defaulted to your desktop
  • baseName: What the file will be called exclusive of a time stamp, defaulted to Output

The script will then take these parameters to create the parameter fileName, which combines the directory location, with what you want the file to be called along with a time stamp, e.g. User\Desktop\Output20170213111157

The command prompt command goes between the parenthesis (). The results of this command will be written to the file.

@ECHO off
REM SET FILE OUTPUT DIRECTORY
SET "filePath=%USERPROFILE%\Desktop\"
REM SET THE BASE OF THE NAME FOR THE OUTPUT FILE 
SET "baseName=Output"
REM CREATE TIMESTAMP
SET timeStamp=%date:~6,8%%date:~3,2%%date:~0,2%%time:~0,2%%time:~3,2%%time:~6,2%
REM CREATE VARIABLE TO HOLD FILENAME WITH A TIMESTAMP
SET "fileNameStamp=%baseName%%Timestamp%"
SET "fileName=%filePath%%fileNameStamp%"
REM CAPTURE COMMAND OUTPUT TO TEXT FILE
> %fileName%.txt (
REM This is where the commands go!
ECHO Text Captured
)

The following working example captures your computer network information and writes it to the file NetworkInfo(time stamp) on your desktop.

@ECHO off
REM SET FILE OUTPUT DIRECTORY
SET "filePath=%USERPROFILE%\Desktop\"
REM SET THE BASE OF THE NAME FOR THE OUTPUT FILE 
SET "baseName=NetworkInfo"
REM CREATE TIMESTAMP
SET timeStamp=%date:~6,8%%date:~3,2%%date:~0,2%%time:~0,2%%time:~3,2%%time:~6,2%
REM CREATE VARIABLE TO HOLD FILENAME WITH A TIMESTAMP
SET "fileNameStamp=%baseName%%Timestamp%"
SET "fileName=%filePath%%fileNameStamp%"
REM CAPTURE COMMAND OUTPUT TO TEXT FILE
> %fileName%.txt (
REM This is where the commands go!
echo Checking your network information, Please wait...

systeminfo | findstr /c:"Host Name" 
systeminfo | findstr /c:"Domain"

ipconfig /all | find "Physical Address" 

ipconfig | find "IPv4" 
ipconfig | find "Default Gateway"
)
Picture showing SQL Server Agent Properties Window

How to setup how much Agent Job History is retained

I was asked to do a historic analysis of SQL Server jobs to determine their statistics, such as the number of runs, average run duration, successes/failures etc. As part of the daily checks I do on a server I’d only check the job runs for the past 24 hours. So I was surprised when I started the analysis and saw that averages for the same job across servers varied significantly.

When I looked at the msdb.dbo.sysjobhistory table of the server with the lowest averages I noted that there were only 999 rows and the history only went back a day or two.

So I had effectively no history on that server. When I looked at another the sysjobhistory table had almost 10 million entries, roughly 4 Gb of data! (The joys of taking over someone else’s environment)

This got me looking at the history settings for the SQL Agent. To open the SQL Server Agent Properties while in SQL Server Object Explorer right click on SQL Server Agent and then click on properties.

Picture showing SQL Server Agent Properties Window

For the first server the history settings were set to the default of 1000 lines for the log as a whole and 100 lines for the individual jobs. If you have a job that runs 4 steps that will use 5 lines, one line for the job and one line for each of the steps that are run. When I looked at the history settings on the second, server as you would expect, no settings had been applied.

Some sensible thresholds needed to be applied across the environment. I’m not going to dictate to you what your retention policy should be regarding job run history but I will say it can be as much as business decision as a server maintenance one so ask around if people need a record that something has run.

You may be in the position that you can be specific enough to enter row values but most organisations will have jobs that run weekly or monthly. In that case you may be better off using the “Older than” option and setting a retention period rather than using rows.

Always with logging, if you have the space, you’re better off having too much than having too little and missing data you need.  The size of each row in the sysjobhistory table is at most 4.5KB so keeping 10k rows is only around 45 Mb. If you don’t have 45 Mb to spare on your server then job history retention is the least of your worries. The worst thing that is likely to happen in storing more rows than you need is that it takes a bit longer to bring up the job history viewer. But here’s a link to some queries to help with that.

The moral of the story is just because it’s in the MSDB doesn’t mean Microsoft are taking care of it. Check your settings and make sure you’re retaining an appropriate amount of history now.

Picture showing the Application screen highlighting the Application type change from Console Application to Windows Forms Application

How to create a console application in Visual Studio that won’t open a command window when it runs

If you’ve written a small console application, say something that doesn’t require a UI e.g. a  batch script that you want to make into an exe, you’ve probably noticed that an unnecessary/unwanted command window pops up.

To stop the command window from opening you can simply make a change in the project properties by setting the Application Type from Console Application to Windows Forms Application.

This can be done from within the project, click on Projects at the top of the screen and then click on Properties as show below.

Picture showing Project options highlighting Properties

Then change the application type from Console Application to Windows Forms Application as shown below.

Picture showing the Application screen highlighting the Application type change from Console Application to Windows Forms Application

Icon of a folder or file with a clock on it

How to get the date of creation and size of current temporary objects from tempdb

The query below will return the temporary objects that currently exist in the tempdb of an instance.

It’s a good query if you’re experiencing issues with the tempdb and want to see what objects are taking up space or if temp tables are surviving longer than you might expect them to.

Recently I found that because queries were running directly from Tableau and the connection was being held open temp tables were not being removed. All the temp tables existing at the same time was taking up several GBs of space needlessly.

-- Get current temporary objects date of creation and size 
SELECT DISTINCT obj.name AS ObjectName
	,obj.type_desc AS ObjectType
	,obj.object_id AS ObjectId
	,obj.principal_id AS PrincipalId
	,obj.schema_id AS SchemaId
	,obj.parent_object_id AS ParentId
	,stat.row_count AS RowCountStat
	,stat.used_page_count * 8 AS UsedSizeKB
	,stat.reserved_page_count * 8 AS RevervedSizeKB
	,obj.create_date AS CreatedDate
	,obj.modify_date AS ModifiedDate
FROM tempdb.sys.partitions AS part WITH (NOLOCK)
INNER JOIN tempdb.sys.dm_db_partition_stats AS stat WITH (NOLOCK) ON part.partition_id = stat.partition_id
	AND part.partition_number = stat.partition_number
INNER JOIN tempdb.sys.tables AS tbl WITH (NOLOCK) ON stat.object_id = tbl.object_id
LEFT JOIN tempdb.sys.objects AS obj WITH (NOLOCK) ON tbl.name = obj.name
ORDER BY CreatedDate ASC;

How to identify databases with Guest user enabled

As best practice it is recommended to disable guest user in every user database, i.e. not master, msdb and tempdb, to improve the security of SQL Server. Guest user permits access to a database for any logins that are not mapped to a specific database user. The guest user cannot be dropped but it can be disabled by revoking the CONNECT permission.

Use the script below to identify which databases have guest user enabled. 

USE master;
GO

DECLARE @database_name SYSNAME
	,@sqlcmd NVARCHAR(4000)

DECLARE databases_cursor CURSOR
FOR
SELECT NAME
FROM sys.databases
WHERE STATE IN (0)
	AND database_id > 4
ORDER BY NAME

CREATE TABLE #guest_users_enabled (
	database_name SYSNAME
	,user_name SYSNAME
	,permission_name NVARCHAR(128)
	,state_desc NVARCHAR(6)
	)

OPEN databases_cursor;

FETCH NEXT
FROM databases_cursor
INTO @database_name;

WHILE @@FETCH_STATUS = 0
BEGIN
	SET @sqlcmd = N'use ' + @database_name + ';

        insert into #guest_users_enabled

        SELECT ''' + @database_name + ''' as database_name, name,

        permission_name, state_desc

        FROM sys.database_principals dpr

        INNER JOIN sys.database_permissions dpe

        ON dpr.principal_id = dpe.grantee_principal_id

        WHERE name = ''guest'' AND permission_name = ''CONNECT'''

	EXEC sp_executesql @sqlcmd

	FETCH NEXT
	FROM databases_cursor
	INTO @database_name;
END

SELECT database_name
	,user_name
	,permission_name
	,state_desc
FROM #guest_users_enabled
ORDER BY database_name ASC

DROP TABLE #guest_users_enabled

CLOSE databases_cursor;

DEALLOCATE databases_cursor;
GO
Use the script below, changing the placeholder database with the database identified in the script above, to revoke permission for the guest user to connect to that database.
USE [database name];

GO

REVOKE CONNECT FROM GUEST;

GO

How to resize TempDB in SQL Server

SQL Server Tempdb is used to store temporary objects. By default the initial size of the tempdb is too small for a production database engaged in any significant enterprise activities, therefore it should be sized accordingly on setup.

Failure to do so will have a negative performance impact when the database is first put into operation as it will need to grow to a more fitting size. Worse still, every time SQL Server is restarted the  tempdb will be recreated with its initial size. So SQL Server has to initiate autogrowth steps to grow the database file again and performance will be impacted negatively while it grows.

To compare the initial tempdb size to the current size run the script below.

USE master;
GO

SELECT mf.database_id
	,mf.NAME
	,mf.size * 8 / 1024 AS Initial_Size
	,df.size * 8 / 1024 AS Current_Size
FROM sys.master_files mf
/*adding info about current file size*/
INNER JOIN tempdb.sys.database_files df ON mf.NAME = df.NAME
/*filtering for tempdb only*/
WHERE mf.database_id = 2;
GO
You can set the tempdb initial size to the displayed current size if you think it will need to grow to this size again or take it as a simple guide and set the initial size less than its current size and allow it to grow as it needs.
To change the tempdb size run the script below replacing all the values in the placeholders with your specific values, use the example script further down as a guide. The files tempdev and templog are typically what the tempdb files are called in a default installation. (You can run the first script again to confirm success)

USE master;
GO

ALTER DATABASE TempDB MODIFY FILE (
	NAME = [logical file name of the tempdb data file]
	,SIZE = [value] MB
	);
GO

USE master;
GO

ALTER DATABASE TempDB MODIFY FILE (
	NAME = [tempdev]
	,SIZE = 4 MB
	);
GO


ALTER DATABASE TempDB MODIFY FILE (
	NAME = [templog]
	,SIZE = 3 MB
	);
GO

How to identify the location of the SQL Server Error Log file

So you want to find out where the SQL Server Error Log file is located for a specific SQL Server instance?

The Error Log is a great place to start tracking down reasons why SQL Server might be experiencing problems.

The queries below (first one works on SQL Server 2008, second one works on 2016) use the extended stored procedure XP_READERRORLOG, which is typically used to read error logs directly, to return the location of the SQL Server Error Log files. The returned value is then cleaned up so only the directory location is provided. You can then copy and paste the result into your file explorer. 

SQL Server 2008

USE Master;
GO

DECLARE @dir AS VARCHAR(MAX)

-- Command will create the temporary table in tempdb
CREATE TABLE [dbo].[#TmpErrorLog] (
	[LogDate] DATETIME NULL
	,[ProcessInfo] VARCHAR(20) NULL
	,[Text] VARCHAR(MAX) NULL
	);

-- Command will insert the errorlog data into temporary table
INSERT INTO #TmpErrorLog (
	[LogDate]
	,[ProcessInfo]
	,[Text]
	)
EXEC [master].[dbo].xp_readerrorlog 0
	,1
	,N'Logging SQL Server messages in file'
	,NULL
	,NULL
	,N'asc';

-- retrieves the data from temporary table and writes it to a variable to remove everything but directory location
SET @dir = (
		SELECT REPLACE(REPLACE(REPLACE([Text], 'Logging SQL Server messages in file ', ''), '''', ''), 'ERRORLOG', '')
		FROM #TmpErrorLog
		)
SET @dir = LEFT(@dir, LEN(@dir) - 1)

SELECT @dir AS ErrorFileLocation

DROP TABLE #TmpErrorLog

SQL Server 2016

USE Master;
GO

DECLARE @dir AS VARCHAR(MAX)

-- Command will create the temporary table in tempdb
CREATE TABLE [dbo].[#TmpErrorLog] (
	[LogDate] DATETIME NULL
	,[ProcessInfo] VARCHAR(20) NULL
	,[Text] VARCHAR(MAX) NULL
	);

-- Command will insert the errorlog data into temporary table
INSERT INTO #TmpErrorLog (
	[LogDate]
	,[ProcessInfo]
	,[Text]
	)
EXEC [master].[dbo].xp_readerrorlog 0
	,1
	,N'Logging SQL Server messages in file';


-- retrieves the data from temporary table and writes it to a variable to remove everything but directory location
SET @dir = (
		SELECT REPLACE(REPLACE(REPLACE([Text], 'Logging SQL Server messages in file ', ''), '''', ''), 'ERRORLOG', '')
		FROM #TmpErrorLog
		)
SET @dir = LEFT(@dir, LEN(@dir) - 1)

SELECT @dir AS ErrorFileLocation

DROP TABLE #TmpErrorLog

 

Five Database icons arranged in a star with the largest one in the center

How to identify all user objects in the SQL Server Master database

What is the SQL Server Master Database used for?

The master database is used by SQL Server to contain all of the system level information:

  • Logins
  • Linked servers
  • Endpoints
  • Configuration settings
  • Information about the other databases on this instance and the location of their files

If the master database is not present, SQL Server cannot start.

This is way you should always take regular backups of the master database.
as if SQL Server suffers a failure, those changes will be lost and you’ll be in a lot of trouble.

So the master database is needed for SQL Server to work. Logic and objects needed to enable your application should not be tangled up in there so you should always avoid creating objects in the master database. Use a user database instead.

In saying that everyone has made the mistake of not specifying the database they want to USE in a script and by default objects get written to the Master database. Or you might have inherited the responsibility for an application which did not conform to best practices and it uses objects created in Master.

To gain visibility of these objects just run the script below.

SELECT o.object_id AS ObjectId
	,o.NAME AS ObjectName
	,o.type_desc AS ObjectType
	,o.create_date AS CreateDate
	,o.modify_date AS ModifyDate
	,SUM(st.row_count) AS RowCnt
	,CAST(SUM(st.used_page_count) / 128.0 AS DECIMAL(36, 1)) AS DataSize_MB
FROM master.sys.objects o
LEFT JOIN master.sys.dm_db_partition_stats st ON st.object_id = o.object_id
	AND st.index_id < 2
GROUP BY o.object_id
	,o.NAME
	,o.type_desc
	,o.create_date
	,o.modify_date
	,o.is_ms_shipped
HAVING o.is_ms_shipped = 0
	AND o.NAME <> 'sp_ssis_startup'
	AND o.type_desc NOT LIKE '%CONSTRAINT%'
ORDER BY CAST(SUM(st.used_page_count) / 128.0 AS DECIMAL(36, 1)) DESC
	,RowCnt DESC