Hands-On Microsoft SQL Server 2008 Integration Services part 25 doc

If you choose a single database from the list, the Object option highlights, enabling you to choose tables or views for which you want to update statistics.. However, if you choose multi

Notify Operator Task

The Notify Operator task sends notification messages to SQL Server Agent operators For this task to work properly, you need to have the SQL Server Agent service

working, and for this task to show you the list of operators, you need to have operators defined in the SQL Server Agent and assign them an e-mail address This task lists only users that have an e-mail address associated with them

Figure 5-35 Configuration options for the Maintenance Cleanup task

Following is the step-by-step method to configure this task:

1 Specify an ADO.NET (SqlClient Data Provider) Connection Manager in the

Connection field If you do not have any operator configured with an e-mail

address to which you are trying to connect, you will receive the error message

“There are no operators with e-mail addresses defined on this server.” You can

configure operators using SQL Server Management Studio

2 Select the operators you want to notify from the list, as shown in Figure 5-36

under the Operators To Notify area

3 Type in the subject and the message you want to send in the provided fields and

click OK to complete the configuration

Figure 5-36 User Interface for the Notify Operator task

Rebuild Index Task

When you insert, update, or delete records from tables, the indexes become fragmented and the performance suffers To keep the performance at the optimal level, you may decide

to rebuild indexes as part of your SSIS package The Rebuild Index task drops the index and creates a new one using the specified fill factor for SQL Server database tables and views By rebuilding the index, this task removes fragmentation of pages, compacts the tables, reclaims disk space, and orders the rows in the contiguous pages, thus improving performance This task effectively runs the ALTER INDEX REBUILD T-SQL command Here is the step-by-step method to configure this task:

1 Specify the ADO.NET Connection Manager in the Connection field

2 Choose the databases from the following options: All Databases, System Databases (master, msdb, model), All User Databases (excluding master, msdb, model, and tempdb) or select databases from the provided list If you choose a single database from the list, the Object option highlights, enabling you to choose tables or views for which you want to update statistics However, if you choose multiple databases, you can’t select the tables or views and the Object option is dimmed

3 If you have selected a single database in the Databases option, you can select Table, View, or Tables and Views in the Object option (Figure 5-37) Selecting Table or View enables the Selection option

4 When the Selection option is available, you can either select all objects or select objects from the list

5 Choose how much free space you want to leave in a page This free space is governed by the FILLFACTOR represented as a percentage You can either use the original FILLFACTOR by selecting Reorganize Pages with the default amount of free space or you can type in a value for the free space as a percentage when you select the “Change free space per page percentage to” option Based

on your value of free space, a FILLFACTOR value is calculated For example,

if you specify 20% for the free space per page, the FILLFACTOR value will be calculated as 80% The specified values of 0 and 100 are treated similarly, and the pages are filled to the full capacity

6 When you build the index, the intermediate sort results are generally stored in the same database However, you can also store these sort results in the tempdb By doing this, you can reduce the time required to build an index, when the tempdb

is on a separate set of disks than the database in question Note that this may put extra requirements of space on the tempdb So, when you want to use tempdb to store the intermediate sort results to reduce time, you can select the Sort results in tempdb advanced option

7 Choose the “Keep index online while reindexing” advanced option to let the

queries or updates happen to the underlying table during reindexing

Reorganize Index Task

As mentioned earlier, the insert, update, or delete activities on the tables cause indexes

to become fragmented and the performance suffers To fix this, you can use this task in

your SSIS packages; for example, after the Bulk Insert task you may use this task The

Reorganize Index task reorganizes indexes in SQL Server database tables and views

and does the defragmenting of indexes by reordering leaf-level pages and compaction

This task runs the ALTER INDEX T-SQL command and uses the REORGANIZE WITH

(LOB_COMPACTION = ON) clause in the command

Here is the step-by-step method to configure this task:

1 Specify an ADO.NET Connection Manager in the Connection field

2 Choose the databases from the option list (refer to Figure 5-38): All Databases,

System Databases (master, msdb, model), All User Databases (excluding master,

msdb, model, and tempdb), or select databases from the provided list

Figure 5-37 Configuration options for the Rebuild Index task

If you choose a single database from the list, the Object option highlights, enabling you to choose Table, View, or Tables and Views for which you want to update statistics However, if you choose multiple databases in this option, you can’t select the tables or views and the Object option is dimmed

3 Selecting Table or View highlights the Selection option When the Selection option is available, you can either select all objects or select objects from the list

4 You can choose to compact the large objects The data with the image, text, ntext, varchar(max), nvarchar(max), varbinary(max), or xml data type is considered as the Large object data You can optionally choose to view the T-SQL code

Shrink Database Task

This task runs the DBCC SHRINKDATABASE T-SQL command to shrink the specified database The step-by-step configuration process goes like this:

1 Specify an ADO.NET Connection Manager to create a connection to the server

2 Choose the databases from the option list: All Databases, System Databases (master, msdb, model), All User Databases (excluding master, msdb, model, and tempdb), or select databases from the provided list

3 After selecting the database, specify the size in megabytes beyond what you want this task to execute to shrink the database

4 Specify how much free space you want to leave in the database by typing in a value as a percentage (from 0 through 100) in the “Amount of free space to remain after shrink” option (see Figure 5-39) This percentage value is calculated

on the space occupied by the data in the database and not the total size of the database For example, when you have a database of 100MB with 40MB of data

Figure 5-38 Interface for the Reorganize Index task

and 60MB of free space, specifying a 50 percent value in the “Amount of free

space to remain after shrink” option will shrink the database to 60MB, having

40MB of data and 20MB of free space

5 Finally, choose between the two options either to retain the freed space in the

database or to return the freed space to the operating system after shrinking

Update Statistics Task

This task runs UPDATE STATISTICS T-SQL command to update information about

the distribution of key values for one or more statistics groups (collections) in the

specified databases Following is the step-by-step procedure for configuring this task:

1 Specify an ADO.NET (SqlClient Data Provider) Connection Manager to connect

to the server

2 Choose one or more databases from the option list: All Databases, System

Databases (master, msdb, model), All User Databases (excluding master, msdb,

model, and tempdb), or select databases from the provided list

3 If you choose a single database from the list, the Object option highlights,

enabling you to choose Table, View, or Tables and Views for which you want to

update statistics (see Figure 5-40) However, if you choose multiple databases in

this option, you can’t select the tables or views and the Object option is dimmed

4 Selecting Table or View highlights the Selection option When the Selection

option is available, you can select all objects or select objects from the list

Figure 5-39 Interface for the Shrink Database task

5 Choose between the three options to update All Existing Statistics or Column Statistics Only or Index Statistics Only If you view the T-SQL code, you will see that the Column Statistics Only option adds the WITH COLUMN clause and the Index Statistics Only option adds the WITH INDEX clause to the SQL code

6 Finally choose a full scan or else specify a percentage or number of rows for the scan type

Summary

We have covered a lot of ground in this chapter—all the Control Flow tasks, and we’ve used some of them in Hands-On exercises We have also used variables and property expressions, along with a little exercise on precedence constraints By now you must

be getting ready to create packages involving complex features and data manipulation Before we get into advanced features of event handling and logging (Chapter 8) and data flow (Chapters 9 and 10), we will cover administration of SSIS packages in next two chapters so that while you’re learning about functional objects within SSIS, you also understand the storage and security architecture of Integration Services

Figure 5-40 The Update Statistics task interface

Administering Integration Services

In This Chapter

c Connecting to Integration

Services Service

c Managing SSIS Packages

c Running SSIS Packages

c Summary

When you’re developing a package in Business Intelligence Development

Studio (BIDS), the package is stored in the file system Once the package has been developed, you can choose to save the developed package either

on the file system or inside the MSDB database of SQL Server Later you may use SQL Server Management Studio to run, monitor, or manage the packages from the chosen storage location SQL Server Management Studio provides these management capabilities using a Windows service—the SQL Server Integration Services 10.0 service Using this service, you can manage SSIS packages on local or remote servers Integration Services service lets you stop and start SSIS packages, monitor running packages, import and export packages, and manage and customize the package store SQL Server Management Studio provides a GUI interface to manage SSIS packages when you connect to Integration Services Another utility, dtutil, helps you to manage SSIS packages from a command line It is important to understand that the Integration Services service is a Windows service for managing SSIS packages and behaves in the same way as other Windows services do; for instance, you can use tools to connect

to it and get your requests serviced such as to copy or run a package SQL Server Management Studio is the out-of-the-box application that allows you to connect to the Integration Services service and manage the stored and running SSIS packages, although you can also use the dtutil utility or a custom-built application to do the same During installation, when you choose Integration Services in the Feature Selection page, SQL Server Integration Services 10.0 service is installed If you don’t install the Integration Services service, you cannot manage your packages using SQL Server Management Studio; however, you can still save and execute your packages while working in BIDS Another benefit of having the Integration Services service installed is that it caches the SSIS tasks and components metadata when the package is loaded the first time and speeds up subsequent loading of the package

Connecting to Integration Services Service

As mentioned earlier, you can connect to Integration Services using SQL Server Management Studio When you install the Integration Services service, by default, it

is started and is ready to be connected to Generally, you’ll be connecting to it with the default settings; however, you may have reasons to connect differently You will be using Integration Services to manage your packages in the following scenarios:

Managing packages with default settings c

Managing packages saved on a remote server c

Managing packages on an instance of SQL Server c

Connecting to Integration Services on a remote server c

Managing Packages with Default Settings

Integration Services creates a configuration file, MsDtsSrvr.ini.xml, on installation The

default configuration file for Integration Services is located in the folder C:\Program

Files\Microsoft SQL Server\100\DTS\Binn, though your location may differ, depending

upon your installation Locate the file, right-click it, and choose Open With | Notepad

You will see the following XML configuration code in the file:

<?xml version="1.0" encoding="utf-8"?>

<DtsServiceConfiguration xmlns:xsd="http://www.w3.org/2001/

XMLSchema" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">

<StopExecutingPackagesOnShutdown>true

</StopExecutingPackagesOnShutdown>

<TopLevelFolders>

<Folder xsi:type="SqlServerFolder">

<Name>MSDB</Name>

<ServerName>.</ServerName>

</Folder>

<Folder xsi:type="FileSystemFolder">

<Name>File System</Name>

<StorePath> \Packages</StorePath>

</Folder>

</TopLevelFolders>

</DtsServiceConfiguration>

When you connect to Integration Services using SQL Server Management Studio,

you see the File System and MSDB root-level folders, which are explained in detail

later in this chapter For now, let’s focus on connectivity; note that the ServerName

element in the configuration file contains a dot (.), which is pointing to the default

instance of the local SQL Server This means by default, you connect to the MSDB

database on the default instance of local SQL Server The File System folder is

configured by the FileSystemFolder folder type element, which points to C:\Program

Files\Microsoft SQL Server\100\DTS\Packages folder by default

Managing Packages Saved on a Remote Server

You might have saved packages to SQL Server on a remote server and want to manage

those packages by connecting from your client machine In this case, you will have

SQL Server Management Studio running on your client machine and will connect to

the local instance of Integration Services However, the local instance is configured to

manage the storage areas such as the MSDB database that are on the remote server

to show packages saved to SQL Server and a UNC path for packages saved to File

System The configuration file is modified to accomplish this