Table of Contents

Lab
2-1: Creating Source Definitions
2-2: Creating Target Definitions
2-3: Creating Mappings

1
11
17

3-1: Create and Run Workflows

25

4-1: Using Filters and Expressions

4-2: Features and Techniques

35
45

5-1: Joining Data

51

6-1: Using the Lookup Transformation

65

7-1: Using Aggregators and Sorters

73

8-1: Using the Debug Wizard

81

9-1: Updating Target Tables

91

10-1: Filling In Missing Data

10-2: Using a Persistent Variable

101
109

11-1: Creating a Mapplet

115

12-1: Using Assignment and Decision Tasks

12-2: Using Events and Timers

121
129

Mapping Design Workshop

Workflow Design Workshop

137
145

Lab 2-1: Creating Source Definitions

Scenario:

You will create some Source definitions for use in later work.

Goals:
Use wizards to import Source definitions from a flat file and a relational database table
Preview the data in the Sources
Duration:
10 minutes

Instructions
Note: Throughout this and later exercises, xx will refer to the student number assigned to
you by your Instructor or the machine you are working on. For example, if you are
Student05, then ~Developerxx refers to the folder ~Developer05.

Step 1.

Start PowerCenter Designer

1)

On your desktop, double-click the PowerCenter Designer icon (

2)

In the Repository Navigator, double-click REP_SVC_EDW_DEV.

Figure 1:

3)

) to start it.

Double-click the Repository to open it

In the "Connect to Repository" dialogue:

a)

For Username, enter Devxx (xx is the number assigned by your instructor).

b)

For Password, enter Devxx.

c)

Click Connect.

PowerCenter 8.x Level I Developer

Lab 2-1

Figure 2:

4)

Connect to Repository dialogue

Right-click the folder labeled ~Developerxx and select Open.

Figure 3:

Open the Folder

Note: In future instructions this may be referred to as "your folder."

Lab 2-2

PowerCenter 8.x Level I Developer

Step 2.

Create a Flat File Source

1)

From the menu, select Tools Source Analyzer.

2)

From the menu, select Sources Import from File.

3)

In the Open Flat File dialogue, select customer_central.dat and click OK.
a)

4)

The file is located in C:\infa-shared\Srcfiles

In Step 1 of the Flat File Import Wizard:

a)

Make sure that the Delimited radio button is selected.

Note: This is because you are importing a comma-delimited file. You will select the field
delimiter on the next screen. Note that PowerCenter can also import files with fixed
field widths.
b)

Check Import field names from first line.

(i)

Note that "Start Import at Row" automatically changes to "2".

Note: When a file has column names, as this file does, PowerCenter can import those as field
names.
c)
Figure 4:

Click Next.
Flat File Import Wizard - Step 1 of 3

PowerCenter 8.x Level I Developer

Lab 2-1

5)

In Step 2 of the Flat File Import Wizard:

a)

In the Delimiters section, make sure that Comma is checked.

Note: While a number of standard delimiters are listed, you can define any character or set
of characters as the delimiter using the "Other" checkbox.
b)
Figure 5:

Accept the defaults for other values and click Next.

Flat File Import Wizard - Step 2 of 3

Note: This step sets up the fields in general. You will have the opportunity to adjust
individual fields in Step 3.
Note: "Use default text length" - check this to set a standard length for all fields with a text
data type. Leave it unchecked and PowerCenter derives the text field length from the
actual length of the data in the file.
Note: "Escape Character" is the character used in your file format if the delimiter character
may appear in a field. Consult documentation to learn more.
6)

In Step 3 of the Flat File Import Wizard:

a)

Use the scrollbar to move to the right and select the City field.

b)

Change the length/precision of the City field to 50

c)

Change the length/precision of the State field to 15

Note: These lengths come from data analysis.

d)

Select the ZIP field.

Lab 2-2

PowerCenter 8.x Level I Developer

Note: You will adjust this field because you know in the future you will import addresses
with Canadian postal codes, which contain alphanumeric characters.
e)

Change the field type to Text.

f)

Change the Length/Precision to 6.

g)

Scroll to the far right and select the field DATE.

h)

Change the field type to Datetime.

Note: You will adjust this field because you know that in the target database the data will be
stored as a Date. You could perform the conversion using PowerCenter's "To-Date"
function, but it is simpler to use the implied conversion functionality of the Source
definition.
i)
Figure 6:

7)

Click Finish.
Flat File Import Wizard - Step 3 of 3

The Source definition will appear in the Source Analyzer workspace.

PowerCenter 8.x Level I Developer

Lab 2-1

Figure 7:

8)

Source Definition Created

Edit the Source definition.

Note: You may use this same Source definition to import data from multiple flat files. You
will change the name to refer generically to customers rather than the specific customer
data file.
a)

Double-click the green header bar at the top of the Source definition.
(i)

b)
c)
d)
e)
9)

This opens the definition for editing.

In the Edit Tables dialogue, click Rename.

For Table Name, enter Customers.
Click OK
Click OK.

To verify that the Source imported correctly, you will now preview the contents of
the flat file.
a)

Right-click the header bar of the Source definition and select Preview data.

b)

In the Preview Data dialogue:

Figure 8:

Preview Data dialogue

(i)

Click the button labeled with three dots ().

(ii) In the Open Flat File dialogue, select customer_central.dat and click Open.
(iii) Click Open.
(iv) The Preview Data dialogue will display data from the flat file.

Lab 2-2

PowerCenter 8.x Level I Developer

Figure 9:

Previewing Data in the Flat File

(v) Click Close.

Create a Relational Database Table Source

Step 3.
1)

Right-click on the Source Analyzer workspace and select Clear All.

2)

Select Sources Import from Database.

3)

In the Import Tables dialogue:

a)

For ODBC data source, select OLTP (DataDirect 5.2 Oracle Wire Protocol).

b)

For Username, enter SDBU.

(i)

c)
d)

Owner Name will automatically populate with SDBU

For Password, enter SDBU.

Click Connect.

PowerCenter 8.x Level I Developer

Lab 2-1

Figure 10:

4)

Import Tables dialogue

e)

In the Select tables pane, click the plus sign (+) beside SDBU to expand it.

f)

Expand TABLES similarly.

g)

Select DEALERSHIP.

h)

Click OK.

i)

The source definition DEALERSHIP (Oracle) appears in the Source Analyzer

workspace.

As before, verify the Source by previewing the data.

a)

Right-click the Source and select Preview Data.

b)

In the Preview Data dialogue:

(i)

Select ODBC data source OLTP (DataDirect 5.2 Oracle Wire Protocol).

(ii) For Username, Owner name, and Password, enter SDBU. (Owner name
should populate automatically.)
Figure 11:

Preview Data Dialogue for Relational Source Definition

(iii) Click Connect.

Lab 2-2

PowerCenter 8.x Level I Developer

Figure 12:

Previewing Data in the Relational Source

(iv) Click Close.

Step 4.
1)

Save Your Work

Type Ctrl+S to save your work.

Note: Always save your work before closing the application or moving on to another task.
There is no automatic save in PowerCenter.
Note: You can also save by selecting RepositorySave from the menu.

PowerCenter 8.x Level I Developer

Lab 2-1

10

Lab 2-2

PowerCenter 8.x Level I Developer

Lab 2-2: Creating Target Definitions

Scenario:
You will create a relational Target to receive data from the flatfile Source you created in
Lab 1
Goals:

Create a Target definition from scratch

Create a Target definition from a Source definition and change the Target type

Duration:
10 minutes

Instructions
Step 1.
1)

Define a Target
Determine what columns will be required
a)

Figure 13:

2)

In PowerCenter Designer, drag the DEALERSHIP Source onto the Source

Analyzer workspace and examine it to determine what columns a target based
on this source will require.
DEALERSHIP Source

Create a Target definition.

a)

In PowerCenter Designer, select the Target Designer tool (

b)

From the Designer menu, select TargetsCreate.

PowerCenter 8.x Level I Developer

Lab 2-2

).

11

c)

In the Create Target Table dialogue:

(i)

Enter the name STG_DEALERSHIP

(ii) Select the database type Oracle

Figure 14:

Create Target Table

(iii) Click Create.

(iv) Click Done.
d)
Figure 15:

12

The Target definition appears in the Target Designer workspace, with no

columns.
New Target

Lab 2-2

PowerCenter 8.x Level I Developer

3)

Add a column to the Target definition.

a)

Double-click the header of the STG_DEALERHIP Target definition to open

the Edit Tables dialogue.

b)

Select the Columns tab.

Figure 16:

c)

Columns tab of the Edit Tables dialogue

Click the New Column button (

(i)

) to create a new column.

For the Name, enter DEALERSHIP_ID

(ii) Set the Datatype to number(p,s)

(iii) For the Precision, enter 3
(iv) For Scale, leave the default 0
(v) DEALERSHIP_ID is the unique identifier of a dealership. Therefore, select
Primary Key in the Key Type column.

PowerCenter 8.x Level I Developer

Lab 2-2

13

d)

Repeat this process for the remaining six columns.

(i)

Figure 17:

Completed DEALERSHIP Target Table

e)

When finished, click Apply then OK.

f)

The Target definition should look like this:

Figure 18:

14

Accept the default (Not a Key) for Key Type.

Completed STG_DEALERSHIP Target Definition

Lab 2-2

PowerCenter 8.x Level I Developer

Step 2.
1)

Create a Target Definition from a Source

In the Repository Navigator, locate the Customers flat file Source definition you
created in the previous lab and drag it into the Target Designer workspace.

Figure 19:

2)

Drag the Source Definition into the Target Designer

Double-click the header of the Customers Target definition to open the Edit Tables
dialogue and select the Columns tab.
a)

Figure 20:

Note that the Datatypes are "number" and "string," as is standard for flat file
definitions.
Columns Tab

3)

Select the Table tab.

4)

Click the Rename button and change the target name to STG_CUSTOMERS.

5)

Change the "Database type" dropdown to Oracle.

6)

Click Apply.
a)

The Tables tab should now look like this:

PowerCenter 8.x Level I Developer

Lab 2-2

15

Figure 21:

7)

Return to the Columns tab.

a)

Figure 22:

8)

Note that the "number" and "string" datatypes have changed to the
"number(p,s)" and "varchar" types appropriate to Oracle.
Column Datatypes Changed

Scroll down and locate the field DATE_FLD.

a)

9)

Tables Tab Edited

Change its datatype to "Date."

Click OK

10) Save your work.

16

Lab 2-2

PowerCenter 8.x Level I Developer

Lab 2-3: Creating Mappings

Scenario:
You need to create Mappings to connect Sources directly to Targets (passthrough
Mappings) so that data can be loaded into the staging tables
Goals:

Create shortcuts to objects in the shortcut (SC) folder

Create a pass-through Mapping that brings data from a single Source to a single Target

Duration:
30 minutes

Instructions
Step 1.

Create Shortcuts

Note: Best practices call for developers to build mappings from shortcuts to a common
folder, rather than defining Sources and Targets in the developers' own folders. This has
several advantages, of which the most significant is that it greatly eases migration of
mappings between PowerCenter environments (e.g., from Development to Test to
Production). Developers create sources and targets, and the Administrator copies them
to the Shortcut folder, where they can be used by all developers, and in migration.
In this lab, you will use shortcuts based on the Sources and Targets you created in labs 1 and
2. The administrator has already copied these Sources and Targets. You will learn how to
create shortcuts to objects in the shortcut folder.
Note: Best practices also call for data to be loaded directly from Sources into staging tables as
part of the ETL process. From these tables, data can be accessed for transformation and
loading without putting a further burden on the Source systems.
1)

Make a shortcut to the Customers flatfile Source.

a)
b)
c)

In PowerCenter Designer, open the Source Analyzer workspace.

If any Source definitions are showing, right-click the workspace and select Clear
All
Locate the folder SC_DATA_STRUCTURES.

Note: The SC_ prefix is a Velocity standard for shortcut folders.

d)

Click once on the plus sign (+) to expand the folder.

Note: Do not double-click the name of the folder. This will connect you to the folder, and
you need to remain connected to your own ~Developerxx folder to create the shortcuts.
e)

Click once on the plus sign to the left of the subfolder named Sources.

PowerCenter 8.x Level I Developer

Lab 2-3

17

f)
Figure 23:

Click once more on the plus sign to the left of the FlatFile subfolder. The
Repository Navigator should now look like this:
Repository Navigator - Shortcut Folder

g)

Click the Customers flatfile Source definition and drag it into the Source
Analyzer workspace.

h)

In the Designer dialogue, click Yes.

Figure 24:

Designer dialogue

Note: If the dialogue asks you whether to copy the source table, say No and try again. You
want to make a shortcut, not a copy.
i)

18

Double-click the Shortcut_to_Customers Source definition to edit it.

Lab 2-3

PowerCenter 8.x Level I Developer

j)

Click Rename.

k)

Change the name of the source definition to SC_Customers.

l)

Click OK to close the Edit Tables dialogue.

Note: The SC_ prefix is Velocity best practice for all shortcuts to objects in other folders.
m) In the Repository Navigator window, expand your ~Developerxx folder, then
the Sources sub-folder, then the FlatFile sub-subfolder.
n)

Confirm that the shortcut appears there.

2)

Following the same procedure, create a shortcut to the OLTP Source

DEALERSHIP and rename it SC_DEALERSHIP.

3)

Open the Target Designer.

4)

Following the same procedure, create a shortcut to the Target STG_CUSTOMERS

and rename it SC_STG_CUSTOMERS.

5)

Following the same procedure, create a shortcut to the Target STG_DEALERSHIP

and rename it SC_STG_DEALERSHIP.

6)

Save your work.

Step 2.

Create the First Mapping

1)

Click the Mapping Designer button (

2)

In this step, you will place all required components into the Mapping Designer
workspace.
a)

In your ~Developerxx folder, expand the Sources subfolder, then the OLTP
sub-subfolder.

Figure 25:

b)

) to open the Mapping Designer tool.

Expanding the Subfolders

Drag the SC_DEALERSHIP shortcut onto the Mapping Designer workspace.

(i)

You will be prompted to name the new Mapping. Give it the name
m2_STG_DEALERSHIP_xx. (Do not type "xx" - use your student id
number!)

Note: Velocity best practice is for all Mappings to begin with the identifying prefix "m_"

PowerCenter 8.x Level I Developer

Lab 2-3

19

(ii) Note that both a Source and a Source Qualifier transformation appear in
the Mapping Designer. If they did not, contact your instructor for help.
Figure 26:

c)
d)
3)

SC_DEALERSHIP source and Source Qualifier

In your ~Developerxx folder, expand the Targets subfolder.

Drag the SC_STG_DEALERSHIP shortcut from the Targets folder onto the
Mapping Designer workspace.

In this step, you will link the Source Qualifier to the Target.

Hint: This procedure may be easier if you rearrange the column widths in the Source
Qualifier and Target so that you can see the full name of the port

20

a)

In SQ_SC_DEALERSHIP, click DEALERSHIP_ID and keep your left mouse

button depressed.

b)

Drag the mouse pointer to SC_STG_DEALERSHIP, still keeping the button

depressed.

c)

Hover over DEALERSHIP_ID on the target and release the mouse button.

d)

A blue arrow, representing a link between the ports of the Source Qualifier and
the Target definition, appears.

e)

Repeat this process to link all ports of the Source Qualifier to the similarlynamed ports in the Target.

Lab 2-3

PowerCenter 8.x Level I Developer

f)

The Mapping should look like this:

Figure 27:

4)

Save your work. In the Output Window, verify that the Mapping is valid. (If it is
not, and you cannot spot the error, ask your instructor for help.)

Step 3.
1)

Create the Second Mapping

From the menu, select Mappings Create.
a)

2)

Completed mapping m2_STG_DEALERSHIP_xx

Name the new Mapping m2_STG_CUSTOMERS_xx

Drag the Source definition SC_Customers (the flat file definition, not the OLTP
definition) and the Target definition SC_STG_CUSTOMERS into the Mapping
Designer workspace.
a)
b)

A Source Qualifier, SQ_SC_Customers, also appears.

Using the process you have already learned, link the ports from the Source
Qualifier to the same-named ports in the Target definition.
(i)

The DATE port in the Source Qualifier does not have a same-named port
in the Target definition. Link it to the DATE_FLD port in the Target
definition.

PowerCenter 8.x Level I Developer

Lab 2-3

21

Figure 28:

3)

22

Completed mapping m2_STG_CUSTOMERS_xx

Save your work and verify that the mapping is valid.

Lab 2-3

PowerCenter 8.x Level I Developer

Step 4.

Create More Shortcuts

In the remaining labs of this class, you will use shortcuts to the objects in
SC_DATA_STRUCTURES. It will be convenient to create those shortcuts now so they
will be available in later labs.
1)

In PowerCenter Designer, open the Source Analyzer.

2)

Drag every Source definition in the SC_DATA_STRUCTURES folder onto the

Source Analyzer workspace, except for Customers and DEALERSHIP (because you
already have shortcuts to those).

3)

Open the Target Designer.

4)

Drag every Target definition in the SC_DATA_STRUCTURES folder onto the

Target Designer workspace, except for STG_CUSTOMERS and
STG_DEALERSHIP (because you already have shortcuts to those).
a)

Be careful to create shortcuts, not copies!

Note: The best practice is to change the names of each of these shortcuts to read "SC_"
rather than "Shortcut_to_". However, it is time-consuming and dull.
The labs instructions from here on out will refer to these shortcuts using the "SC_" prefix, to
reflect the best practice; if you do not change the names, simply substitute
"Shortcut_to_".

PowerCenter 8.x Level I Developer

Lab 2-3

23

24

Lab 2-3

PowerCenter 8.x Level I Developer

Lab 3-1: Create and Run Workflows

Scenario:

You need to load the Customer and Dealership data into the Staging tables.

Goals:

Create and run Workflows that execute the Mappings you created in Lab 2-3

Duration:
45 minutes

Instructions
Step 1.
1)

Open Workflow Manager

In PowerCenter Designer, locate the "Tools" toolbar.

Figure 29:

2)

Tools Toolbar

Click the "W" icon to open the Workflow Manager application.

Step 2.

Create a Workflow

1)

If necessary, connect to your ~Developerxx folder.

2)

From the Workflow Manager menu, select Tools Workflow Designer (NOT
Workflow Manager).

3)

From the Workflow Manager menu, select Workflows Create.

4)

In the Create Workflow dialogue

a)

Name the Workflow wkf_STAGE_DEALERSHIP_xx

Note: Velocity best practice is to prefix the name of a Workflow with "wkf_" and give it a
name describing what it does. In this case, we are loading dealership data into a staging
table.
b)
c)

Click OK.
Note that the Workflow is created with a Start task already present.

PowerCenter 8.x Level I Developer

Lab 3-1

25

5)

In the toolbar, click the Session button (

).

a)

Click in the Workflow Designer workspace somewhere to the right of the Start
task.

b)

In the Mappings window, select m2_STG_DEALERSHIPS_xx and click OK.

Figure 30:

c)
Figure 31:

Select Mapping

The Session task is added to the Workflow:

Session Added

Note: The Velocity standard name for a Session task is s_ followed by the name of the
Mapping. The Workflow Designer automatically assigns this name to a Session task
when you add it to the Workflow.

26

Lab 3-1

PowerCenter 8.x Level I Developer

6)

From the Workflow Manager menu, select TasksLink Task

a)

Place the mouse curser over the Start task.

b)

Click and hold the left mouse button.

c)

Drag the cursor to the Session task.

d)

Release the mouse button.

e)

The tasks will be linked:

Figure 32:

7)

Tasks Linked

Double-click the Session task to edit its properties.

Note: The Session task properties determine what files or database tables the Mapping reads
from and writes to. The Source and Target definitions in the Mapping define the fields
to be read or written, but do not directly determine where to read from or write to.

PowerCenter 8.x Level I Developer

Lab 3-1

27

a)

Select the Mapping tab of the Edit Tasks window.

b)

In the navigator on the left, select the Source SQ_SC_DEALERSHIP.

c)

On the right:
(i)

Ensure that the "Reader" selected is Relational Reader.

(ii) In the "Connections" section, click the dropdown arrow and select the
connection SDBU.
Figure 33:

d)
e)

Source Properties

In the navigator on the left, select the Target SC_STG_DEALERSHIP.

On the right:
(i)

Ensure that the "Writer" selected is Relational Writer.

(ii) Ensure that the Relational connection selected is STGxx.

(iii) In the "Properties" section, set the "Target load type" to Normal.
(iv) Scroll down and check "Truncate target table option".
(v) Click OK.
8)

28

Save your work.

Lab 3-1

PowerCenter 8.x Level I Developer

Step 3.

Start the Workflow.

1)

In the "Tools" toolbar, click the "M" icon to start the Workflow Monitor
application.

2)

If the status for "INT_SVC_EDW_DEV" reads "Disconnected", right-click it and

select Connect.

3)

Return to the Workflow Manager.

4)

Right-click in the Workflow Designer workspace and select Start Workflow.

5)

Return to the Workflow Monitor. The status of your Workflow and Session are
now both "Running."

Figure 34:

6)

Workflow Monitor: Workflow Running

When the status changes to "Succeeded" your Workflow has run.

Step 4.

View the Results

1)

Right-click the Session object and select "Get Run Properties."

2)

Click the arrow beside "Source/Target Statistics" to expand it.

Figure 35:

Run Properties and Statistics

PowerCenter 8.x Level I Developer

Lab 3-1

29

3)

Return to the PowerCenter Designer application.

4)

In the m_STG_DEALERSHIPS_xx Mapping, right-click the Target object and

select Preview Data.

5)

In the Preview Data dialogue:

a)

For ODBC Data Source, select STG (DataDirect 5.2 Oracle Wire Protocol).

b)

For Username, Table Owner, and Password, enter STGxx

c)
d)
Figure 36:

30

Click Connect.
The lower part of the Preview Data dialogue populates with the data you loaded
into the STG_DEALERSHIP table.
Preview STG_DEALERSHIP data

Lab 3-1

PowerCenter 8.x Level I Developer

Step 5.
1)

Create and Run a Second Workflow

Following the same procedure, create a second Workflow.
a)

Name the Workflow wkf_STAGE_CUSTOMERS_xx.

b)

Create a Session task using the Mapping m2_STG_CUSTOMERS_xx.

c)
2)

Link the Start task to the Session task.

Double-click the Session task to edit it.

a)

Select the Mapping tab.

b)

Select the source SQ_SC_Customers

(i)

Make sure that the reader type is "File Reader"

(ii) Make sure that the Source Filetype is "Direct"

(iii) Make sure that the Source file directory is "$PMSourceFileDir\"
(iv) Make sure that the Source filename is "customer_central.dat"
c)

Select the target SC_STG_CUSTOMERS

(i)

Make sure that the Writer type is "Relational Writer"

(ii) Make sure that the Connection is "STGxx"

(iii) Make sure that the Target load type is "Normal"
(iv) Make sure that the "Truncate target table option" is checked.
d)

Click OK.

3)

Save your work and make sure the Workflow is valid.

4)

Run the Workflow and review the results.

Figure 37:

Session Details and Statistics for Customer Load

PowerCenter 8.x Level I Developer

Lab 3-1

31

Figure 38:

32

Data Preview of Customers Staging Table

Lab 3-1

PowerCenter 8.x Level I Developer

Extra Credit Lab: File Lists

You may have noted that the name of the Customers file you loaded was
"customer_central.dat". There are two other files, "customer_east.dat" and
"customer_west.dat"
You could load each of these files individually, but there is an easier way: use a file list.
PowerCenter can take as its input a text file listing other files, and will read each of the files
in that list.
Note that PowerCenter will read all the files in the file list using the same Source definition.
This means that the files must have exactly the same format, or PowerCenter will read bad
data.
To make this work, you must make the following changes to the Session object:
The Source Filetype has to be set to "indirect," indicating that the Source file is a file list
rather than a data file.
The Source Filename must be changed to the name of the file containing the file list,
which is "customer_list.dat"

PowerCenter 8.x Level I Developer

Lab 3-1

33

34

Lab 3-1

PowerCenter 8.x Level I Developer

Lab 4-1: Using Filters and Expressions

Scenario:
The customer contacts have been loaded to the Staging table and are ready to be loaded
to ODS
There are bad customer IDs that should not be loaded to ODS
The customer data needs to be reformatted to match the table specification of the ODS
Goals:

Move data from the Customer staging table to the ODS database
o
Use an Expression transformation to reformat data
o
Use a Filter transformation to pass only valid records

Duration:
60 minutes

Instructions
Step 1.

Create the mapping.

1)

Create a new mapping called m4_ ODS_CUSTOMERS_xx.

2)

Add the Source definition SC_STG_CUSTOMERS to the mapping.

3)

Add the Target definition SC_ODS_CUSTOMERS to the mapping.

4)

The Mapping should look like this:

Figure 39:

Mapping Created

PowerCenter 8.x Level I Developer

Lab 4-1

35

Step 2.

Add a Filter Transformation

In this step you will add a Filter transformation to the mapping to pass only records with
valid Customer IDs.
1)

Locate the Filter transformation (

Mapping Designer workspace.

) on the toolbar. Click it and then click on the

2)

Drag the following ports from the Source Qualifier to the Filter transformation:
DEALERSHIP_ID
CUSTOMER_NO
FIRSTNAME
LASTNAME
ZIP
GENDER
INCOME
AGE
DATE_FLD

Figure 40:

36

Ports Connected

3)

Double-click the header of the Filter transformation to edit it.

4)

Rename the filter object fil_Valid_Customer_Number.

Lab 4-1

PowerCenter 8.x Level I Developer

5)

Select the Properties tab.

a)

Click the bent arrow (

Condition.

b)

In the Expression Editor:

(i)

) in the Value column for the attribute Filter

Delete the current expression ("TRUE") from the Expression Editor.

(ii) Enter the expression NOT (ISNULL(CUSTOMER_NO)) AND

CUSTOMER_NO != 99999
(iii) Click Validate
(iv) If the expression is invalid, fix it.
(v) Click OK.
Figure 41:

c)
Figure 42:

Step 3.

Filter Expression

Click OK.
Mapping with Filter Transformation Added

Add an Expression Transformation

In this step you will add an Expression transformation that will format Customer data
correctly for the ODS database.
1)

Locate the Expression transformation (

on the Mapping Designer workspace.

PowerCenter 8.x Level I Developer

Lab 4-1

) on the toolbar. Click it and then click

37

2)

3)

Drag the following ports from the Filter transformation to the Expression
transformation:
FIRSTNAME
LASTNAME
GENDER
INCOME
AGE
Edit the Expression transformation
a)

Rename the Expression transformation exp_Format_Customers.

b)

Select the Ports tab.

c)

Set all the ports you dragged from the Filter to be input-only by unchecking the
"O" column.

d)

Use the New button (

output only:
Port Name
NAME
GENDER_CATEGORY
SENIOR_FLAG
HIGH_INCOME_FLAG

e)
Figure 43:

f)

) to create the following new ports and set them to

Datatype
string
string
integer
integer

Length/Precision
40
7
10
10

The ports should now look like this:

Ports in Expression

Use the Expression Editor to create an expression for the NAME port to
concatenate the FIRSTNAME and LASTNAME fields, with a space in
between:
FIRSTNAME || ' ' || LASTNAME

Note: More advanced data integration developers may recognize that the above expression
leaves something to be desired when dealing with less-than-ideal data, as would be
typical in these fields. Informatica has extensive data quality capabilities to recognize,
cleanse, and supplement name data. These capabilities are in the Data Quality product,
which is outside the scope of this class.

38

Lab 4-1

PowerCenter 8.x Level I Developer

g)

Create an expression for the GENDER_CATEGORY port that expands single

character designations from the GENDER field into full-word descriptions:
DECODE(GENDER, 'M', 'MALE', 'F', 'FEMALE', 'UNK')

The DECODE function uses a mapping to replace the values in a field with other values. In
this case, if the field has a value of "M", then it is changed to "MALE." If the field has a
value of "F", it is changed to "FEMALE." Any other value will be replaced with "UNK"
(for "UNKNOWN"). DECODE is useful when there are a relatively small number of
enumerated values in a field. If there are a larger number of values to be remapped, a
Lookup transformation would be used. (We will cover Lookup transformations later in
this course.)
h)

Create an expression for the SENIOR_FLAG port that sets the port value to 1
(Boolean TRUE) if the AGE is greater than 55:
IIF(AGE > 55, 1)

IIF - "Immediate If" - is a powerful function. When the expression (AGE > 55) evaluates to
TRUE, the first argument is assigned to the port. When the expression does not evaluate
to TRUE, the second argument is assigned to the port. In this case, no second value is
assigned, so the port is set to zero (0) when the expression evaluates to FALSE.
IIF expressions can be nested to handle multibranch logic.
i)

Create an expression for the HIGH_INCOME_FLAG port that sets the port to
1 (Boolean TRUE) if the INCOME is greater than 50000:
IIF(INCOME > 50000, 1)

j)

The ports should now look like this:

Figure 44:

4)

Ports with Expressions Completed

Click OK.

PowerCenter 8.x Level I Developer

Lab 4-1

39

Step 4.
1)

Connect the Ports to Complete the Mapping

Connect the following ports:

From
Transformation
Name

From Port Name

To Transformation
Name

To Port Name

fil_Valid_Customer_
Number

DEALERSHIP_ID

SC_ODS_
CUSTOMER

DEALERSHIP_ID

fil_Valid_Customer_
Number

CUSTOMER_NO

SC_ODS_
CUSTOMER

CUSTOMER_NO

fil_Valid_Customer_
Number

ZIP

SC_ODS_
CUSTOMER

POSTAL_CODE

fil_Valid_Customer_
Number

DATE_FLD

SC_ODS_
CUSTOMER

CONTACT_DATE

exp_Format_
Customers

NAME

SC_ODS_
CUSTOMER

NAME

exp_Format_
Customers

GENDER_
CATEGORY

SC_ODS_
CUSTOMER

GENDER_
CATEGORY

exp_Format_
Customers

SENIOR_FLAG

SC_ODS_
CUSTOMER

SENIOR_FLAG

exp_Format_
Customers

HIGH_INCOME_
FLAG

SC_ODS_
CUSTOMER

HIGH_INCOME_
FLAG

Figure 45:

2)

Save your work.

a)

40

Target, Connected

Verify that the Mapping is valid and fix any problems that keep it from
validating.

Lab 4-1

PowerCenter 8.x Level I Developer

3)

The Mapping should look like this:

Figure 46:

Step 5.

Completed Mapping

Create and Run the Workflow

1)

Open PowerCenter Workflow Manager.

2)

Create a new workflow named wkf_Load_ODS_CUSTOMERS_xx.

3)

Add a Session task using the mapping m4_ODS_CUSTOMERS_xx.

4)

Edit the Session task and set the connection values:

a)

Click the Mapping tab

b)

Source connection should be STGxx

c)

Target connection should be ODSxx

d)

Make sure to set the Target load type to Normal

e)

Make sure to set the Truncate target table option

5)

Link the Session to task the Start task

6)

Save your work.

7)

Start the workflow and monitor the results.

Figure 47:

Results of Running the Workflow

PowerCenter 8.x Level I Developer

Lab 4-1

41

Figure 48:

Data Preview of ODS_CUSTOMERS

Note: The number of rows you see may differ from what is shown in the figure, depending
on whether you performed the extra credit exercise at the end of Lab 4.
a)

42

Why does the number of rows in the source not match those in the target?

Lab 4-1

PowerCenter 8.x Level I Developer

PowerCenter 8.x Level I Developer

Lab 4-1

43

Answers
5.7.a. Why does the number of rows in the source not match those in the target?
Some rows were removed by the Filter transformation, so those rows did not reach the target.

44

Lab 4-1

PowerCenter 8.x Level I Developer

Lab 4-2: Features and Techniques

Goals:
In this lab you will learn and practice some features and techniques that will increase
your efficiency as a PowerCenter Developer.
Duration:
30 minutes

Instructions
WARNING: In this lab, do not save your work. While it is normally best practice to save
your work frequently while working in PowerCenter, in this case you will be making
changes to a Mapping that is already the way you want it. So don't save your work!

Step 1.

Arrange All and Arrange All Iconic

In a complex Mapping, it can be hard to see how the parts relate. How can you make this
better?
1)

Begin with the Mapping from Lab 4-1 (m4_ODS_Customers_xx) open in the
PowerCenter Developer application.

2)

Right-click anywhere in the workspace and select Arrange All.

3)

Observe the results.

Arrange All is a tool for arranging the transformations in a Mapping neatly.

4)

Right-click again and select Arrange All Iconic.

Arrange All Iconic enables you to quickly see the relationships between the objects in a
Mapping.

Step 2.
1)

Autolink
"Arrange All" on the Mapping.

2)

Drag the cursor across the links between the Source definition and the Source
Qualifier to select them.

3)

Hit the Delete key on your keyboard.

4)

Right-click and select Autolink by Name.

5)

Position the cursor over the Source, then click and drag to the Source Qualifier.

6)

Click the SQ again to return to the normal (arrow) cursor.

Autolinking provides a quick way to connect the output ports in one transformation to the
input ports in another transformation.
Autolink by Name searches for ports with identical names and connects them
Autolink by Position connects the first output port to the first input port, the second output
port to the second input port, etc.

PowerCenter 8.x Level I Developer

Lab 4-2

45

7)

Delete the links again and Autolink the two by Position.

Step 3.

Select Link Path

Suppose another developer has created a large, complex Mapping that is not working quite
right: some data is winding up in the wrong fields. And you have been asked to debug it.
How can you figure out where the data is coming from? Answer: By tracing the link
paths.
a)

On the Target definition, right-click the POSTAL_CODE field and select

Select Link Path Backward.
(i)

b)

Expand the Filter transformation so you can see the related field there.
(i)

c)

The link to the Postal Code field is now red.

Note that the links leading both into and out of it are red.

You can, by expanding the appropriate transformations, trace the lineage of the
Postal Code field all the way back to the ZIP field in the Source definition.

Selecting the link path enables you to easily trace the lineage of any field forward and
backward through a Mapping.

Step 4.

Propagating Port Properties

You have to change the datatype of a field in the Source. Do you really have to manually
adjust every port along its link path? No.
1)

Edit the Source Qualifier and select the Ports tab.

2)

Change the name of the CUSTOMER_NO port to CUST_NO and its precision
from 5 to 10.

3)

Click OK.

4)

Right-click CUST_NO in the Source Qualifier and select Propagate Attributes.

5)

In the "Propagate Port Attributes" dialogue:

a)

Under "Attributes to Propagate" select Name and Precision, with a direction of

Both.

b)

Click Preview.

c)
d)
e)

f)

46

Note the green and red arrows. What will be changed?

Click Propagate, then Close.
Was a change made in the Filter? What was it?

Was a change made in the Target definition? Why or why not?

Lab 4-2

PowerCenter 8.x Level I Developer

Step 5.

Moving Ports

Sometimes just rearranging the ports on a transformation will make the Mapping easier to
read.
1)

Edit the Filter transformation and select the Ports tab.

2)

Click the AGE port and use the "up arrow" (

list of ports.

3)

Single-click and hold the number next to the ZIP field. Note the square that appears
in the cursor.

4)

Drag ZIP right below AGE.

5)

Click Cancel to discard the changes.

Step 6.

) button to move it to the top of the

Another Method of Creating Transformations

Plus, it bypasses the default names PowerCenter gives a transformation.

1)

From the menu, select TransformationCreate

2)

Select Aggregator from the dropdown box.

3)

Name the Aggregator agg_Demo_Create.

4)

Click Create.
a)

The new transformation appears in the workspace.

5)

By the same method, create a Filter named fil_Demo_Create.

6)

Click Done.

While we're at it, how do you remove an unwanted transformation?

7)

The Filter you just created is already selected. Hold down the Shift key and click the
Aggregator you created to select it, too.

8)

Hit the Delete key on your keyboard.

Step 7.

a)

Note that the Designer dialogue tells you which transformations will be deleted.

b)

Click Yes.

Reverting to Saved

Sometimes you make a mistake that you can't easily undo and need to go back to where you
were before. If you haven't saved, you can do it.
1)

In the Repository Navigator, right-click your folder and select Disconnect.

2)

When asked whether to save the changes to your folder, click No.

3)

Reopen your folder.

4)

If necessary, reopen the Mapping.

5)

Note that it is back to the way it was before Step 1.

6)

Arrange All for the next step.

PowerCenter 8.x Level I Developer

Lab 4-2

47

Step 8.

Scaling

You may not be able to see the whole Mapping in your workspace. But you can.
1)

Maximize PowerCenter Developer.

2)

How many transformations can you see?

3)

In the Standard toolbar at the top of the window, click the Zoom dropbox (
and select 60.

4)

Can you see more transformations?

5)

Click the Scale to fit icon (

6)

You can now see all your transformations at once.

Step 9.

) beside the Zoom dropbox.

Switching Transformations While Editing

When editing several transformations, you don't have to close the Edit Transformations
dialogue and reopen it repeatedly
1)

Double-click the Source Qualifier transformation to edit it.

2)

Select the Ports tab.

3)

In the Select transformation: dropbox, select the Filter transformation.

4)

What happens?

Step 10.

Copy Objects Within and Between Mappings

You may find that you want to duplicate a set of transformations within a Mapping or a
Mapplet, preserving the dataflow between them. This technique may prove useful if you
know that you will need to use the logic contained in the transformations in other
Mappings or Mapplets.

48

1)

Arrange All Iconic.

2)

Use your left mouse button to draw a rectangle that encloses the Filter and
Expression transformations. This will select these objects.

3)

Press Ctrl+C on your keyboard, immediately followed by Ctrl+V.

4)

Note that both transformations have been copied onto the mapping, including the
dataflow between them. They have been renamed with a "1" on the end of their
names.

5)

Open another Mapping.

6)

Press Ctrl+V again.

7)

The transformations are added to the open Mapping.

8)

Disconnect from your folder but do not save the changes (revert to the previously
saved version).

Lab 4-2

PowerCenter 8.x Level I Developer

Step 11.

View Object Dependencies

By viewing object dependencies in the Designer, a user can learn which objects may be
affected by making changes to Source or Target definitions, Mappings, Mapplets, or
transformations. Direct and indirect dependencies are shown.
1)

In the Repository Navigator, select the flat-file Source definition SC_Customers.

2)

Right-click and select Dependencies.

3)

Click OK to show all dependencies.

4)

You will see the View Dependencies window, which will show every Mapping,
Session, and Workflow that uses or depends upon the SC_Customers Source, as
well as those that it uses or depends on.

5)

It also shows the Customers flat file definition in SC_DATA_STRUCTURES that

SC_Customers depends on.

Note: The Save to File button on the View Dependencies window saves the dependency
information as a HTML file (.htm) for later viewing.
6)

Experiment by viewing the dependencies of other objects.

PowerCenter 8.x Level I Developer

Lab 4-2

49

Answers
4.5.e. Was there a change made in the Filter? What was it?
Yes, the name and precision of the Customer Number port changed to match the changes in the Source
Qualifier.
4.5.f. Was there a change made in the Target definition? Why or why not?
No, the Source and Target definitions cannot be changed or edited in the Mapping Designer workspace.
They can only in the Source Analyzer and Target Designer workspaces.
9.4. What happens?
You now see and can work with the ports of the Filter transformation.

50

Lab 4-2

PowerCenter 8.x Level I Developer

Lab 5-1: Joining Data

Scenario:
The ODS requires a Personnel table containing information about employees and the
dealership each works at.
o
Personnel information is stored in two staging tables named STG_EMPLOYEES
and STG_DEALERSHIP
The ODS also requires a Stock Units table containing information about how many and
what kind of units are in stock at the dealerships
o
The Stock Units information is contained in two sources as well, STG_PRODUCT
and a flatfile named Inventory.
Goals:

Create a Mapping with two separate data flows.

Join STG_EMPLOYEES and STG_DEALERSHIP using a single Source Qualifier. Set
the join condition for the tables.
Make a previously-created transformation reusable, and use it in this Mapping.
Join Inventory and STG_PRODUCT using a PowerCenter Joiner transformation.
Create and execute a Workflow to populate both ODS_Personnel and
ODS_StockUnits.
Duration:
90 minutes

Instructions
Step 1.
1)

Create a Homogeneous Join

Create a new Mapping called
m5_Load_ODS_PERSONNEL_STOCK_UNITS_xx

Note: PowerCenter has many options that customize the appearance and functionality of the
client applications. In this case, we want to turn off automatic creation of Source
Qualifiers so we can use a single SQ to create a homogeneous join of two Source
definitions.
2)

From the menu, select Tools Options.

3)

In the Options dialogue:

a)

Select the Tables tab.

b)

Make sure that the Mapping Designer tool is selected.

c)

Uncheck "Create Source Qualifier when opening Sources."

PowerCenter 8.x Level I Developer

Lab 5-1

51

d)
Figure 49:

e)

The Options dialogue should now look like this:

Options Dialogue

Click OK.

4)

Drag the relational table Sources SC_STG_EMPLOYEES and

SC_STG_DEALERSHIP from the Sources - STG folder onto the Mapping
Designer workspace.

5)

In the following steps, you will create a Source Qualifier to join the tables using the
common field DEALERSHIP_ID.

Tip: Note that the fields are of the same data type - if they were not, you could not join the
tables with a single Source Qualifier.
Performance Note: Extensive discussion can ensue when deciding whether it is better to
have the tables joined in the database or by PowerCenter. In general, when the tables
have primary keys and indexes, it is better to join them in the database.
When you are joining more than three tables, database optimizers may or may not devise a
plan that leverages keys and indexes to avoid unnecessary full table scans. If a database
SQL plan analysis indicates that the database is engaging in multiple full table scans,
consider using PowerCenter to join at least some of the relational tables together.

52

Lab 5-1

PowerCenter 8.x Level I Developer

a)

On the Designer toolbar, click the Source Qualifier transformation button

(

).

b)

Click again on the Mapping Designer workspace.

c)

In the "Select Sources for Source Qualifier Transformation" dialogue, make sure
that both SC_STG_DEALERSHIP and SC_STG_EMPLOYEES are selected,
then click OK.

d)

A single Source Qualifier will be created, with all fields from both sources
feeding into it.

Figure 50:

Create the Source Qualifier

e)

Double-click the Source Qualifier to edit it.

f)

Rename it SQ_EMPLOYEE_DEALERSHIP.

g)

Select the Properties tab.

h)

In the User Defined Join field, click the bent arrow to open the SQL Editor and
edit the property.

Tip: Do not use the "Sql Query" field for the Join condition. This will cause the workflow to
fail.
i)

Under STG_DEALERSHIP, double-click DEALERSHIP_ID to place it in the

SQL box.

j)

Type an equal sign (=) in the SQL box.

k)

Click the plus sign by STG_EMPLOYEES to expand it and double-click

DEALERSHIP_ID.

PowerCenter 8.x Level I Developer

Lab 5-1

53

l)
Figure 51:

The SQL Editor should now look like this:

SQL for Homogeneous Join

m) Click OK.
n)

Click OK again to close the Edit Transformations dialogue.

o)

Save your work.

Hint: The mapping will not validate, as it does not yet have a target object. This is OK.

Step 2.

Create a Reusable Transformation

1)

Open the Mapping m4_ODS_CUSTOMERS_xx.

2)

Double-click the transformation "exp_Format_Customers" to edit it.

a)

54

Make sure the Transformation tab is selected.

Lab 5-1

PowerCenter 8.x Level I Developer

b)
Figure 52:

Check the option "Make Reusable."

Making a Transformation Reusable

Note: Making a transformation reusable is not reversible. Once done, it cannot be revoked.
Note: Also note that the best practice says that reusable objects should be created in the
project shortcut folder (in this class, SC_DATA_STRUCTURES). If you, as a
developer, promote an object to be reusable, you should notify your tech lead so s/he can
move it to the appropriate shortcut folder. This enables the object to be properly
migrated to the Test and Production environments. Once the tech lead has done this,
you must modify your mapping to use the shortcut rather than the object from your
local folder.
c)
d)
e)
3)

Click Yes to make the transformation reusable.

Select the Ports tab and note that the expressions are read-only (grayed out).
You cannot edit a reusable transformation in the context of a Mapping.
Click OK to close the Edit Transformations dialogue.

Select the Transformation Developer tool (

tool bar).

PowerCenter 8.x Level I Developer

Lab 5-1

). (This can be found on the Tools

55

4)

In your ~Developerxx folder, open the Transformations subfolder and drag

exp_Format_Customers onto the Transformation Developer workspace. Your
screen should look like this:

Figure 53:

5)

Transformation Developer

Double-click the transformation to open the Edit Transformations dialogue.

a)

Select the Ports tab.

b)

Note that the expressions are editable here.

c)
d)

Select the Transformation tab.

Change the name of the transformation to re_exp_Format_Persons to more
accurately reflect its role (formatting both Customer and Personnel data).

Note: Velocity best practice is to prefix re_ to the name of any reusable transformation.

6)

Step 3.

56

e)

Do not change anything else.

f)

Click OK.

Save your work.

Complete the First Data Flow

1)

Open the Mapping m7_Load_ODS_Personnel_Stock_Units_xx.

2)

Drag the reusable expression re_exp_Format_Persons from the Transformations

sub-folder into the Mapping Designer workspace.

3)

Connect the ports FIRST_NAME and LAST_NAME from the Source Qualifier to
the ports FIRSTNAME and LASTNAME in the Expression transformation.

4)

Connect the port GENDER in the Source Qualifier to the port GENDER in the
Expression transformation.

5)

Connect the port AGE in the Source Qualifier to the port AGE in the Expression
transformation.

6)

Drag the Target SC_ODS_PERSONNEL into the Mapping Designer workspace.

Lab 5-1

PowerCenter 8.x Level I Developer

7)

Connect the Target's ports according to the following table:

Expression

Port

re_exp_Format_Persons

NAME
GENDER_CATEGORY
SENIOR_FLAG
EMPLOYEE_ID
DEALERSHIP_ID
ZIP_CODE
HIRE_DATE
POSITION_TYPE

SQ_Employee_Dealership

DEALERSHIP_MANAGER_ID

8)

DEALERSHIP_DESC
DEALERSHIP_LOCATION
Save your work. Verify that the Mapping is valid.

9)

Arrange All Iconic.

Figure 54:

Step 4.

Port in
SC_ODS_PERSONNEL
NAME
GENDER_CATEGORY
SENIOR_FLAG
EMPLOYEE_ID
DEALERSHIP_ID
POSTAL_CODE
HIRE_DATE
POSITION_TYPE
DEALERSHIP_MANAGER_ID
DEALERSHIP_DESCRIPTION

DEALERSHIP_LOCATION

Personnel Data Flow Arranged Iconic

Create a Heterogeneous Join

1)

Turn on automatic creation of Source Qualifiers (see Figure 1 of this lab).

2)

Add the transformations to the Mapping:

a)

Drag the flatfile Source definition SC_Inventory and the relational Source
definition SC_STG_PRODUCT into the Mapping Designer workspace.

Since you will be joining a flatfile source to a relational source, you cannot use a
homogeneous join here. Therefore you will use a Joiner transformation.
b)
c)

Click the Joiner transformation icon (

) and click again in the workspace.

Drag the Target definition SC_ODS_STOCK_UNITS into the mapping.

PowerCenter 8.x Level I Developer

Lab 5-1

57

d)
Figure 55:

The new portion of the Mapping should look like this:

New Portion of the Mapping

Performance Note: The PowerCenter Joiner transformation is fast and uses RAM rather
than disk memory wherever possible. Optimizing the use of RAM can be important,
particularly when RAM space is limited. Therefore, the Master side of the Joiner should
be the one with the fewest duplicate keys and the fewest rows (provided this fits the logic
of the join). Also, joining sorted data allows more efficient use of RAM.
In this lab, we will adhere to best practice by using STG_PRODUCT as the Master side of
the Joiner. STG_PRODUCT has a much smaller number of rows than Inventory, and
no duplicate keys.
3)

Drag the following ports from SQ_SC_Inventory to the Joiner:

4)

INVENTORY_ID
PRODUCT_ID
DEALERSHIP_ID
RECEIVED_DATE
QTY_ON_HAND
INVOICE_PRICE
TIME_KEY
MSRP
Drag the following ports from SQ_SC_STG_PRODUCT to the Joiner:

5)

PRODUCT_ID
GROUP_ID
PRODUCT_DESC
GROUP_DESC
DIVISION_DESC
Double-click the Joiner transformation to edit it.
a)

Rename it jnr_Inventory_FF_STG_PRODUCT.

Note: the FF is for "Flat File." As a general rule, naming conventions should be as clear as
possible.
b)

Select the Ports tab.

c)

Change the name of the port PRODUCT_ID to

PRODUCT_ID_INVENTORY.

d)

Change the name of the port PRODUCT_ID1 to

PRODUCT_ID_PRODUCT.

e)

58

Select the Condition tab.

Lab 5-1

PowerCenter 8.x Level I Developer

f)

Create a new Join condition.

g)

Make sure that the value in the Master column is PRODUCT_ID_PRODUCT

and the value in the Operator column is =.

h)

Change the value in the Detail column to PRODUCT_ID_INVENTORY.

i)

The Condition tab should look like this:

Figure 56:

j)

Step 5.

Condition for jnr_Inventory_FF_STG_PRODUCT

Click OK.

Filter Products Not In Inventory

In this step you will create a Filter transformation to remove products with no inventory.
1)

Add a new Filter transformation and call it fil_Available_Stock.

2)

Drag every port except for PRODUCT_ID_PRODUCT (which is redundant) from

the Joiner transformation to the Filter transformation.

3)

Set the Filter condition so that QTY_ON_HAND must be greater than 0.

4)

Change the name of the PRODUCT_ID_INVENTORY port to PRODUCT_ID.

5)

Use Autolink by Name to connect ports from the Filter transformation to the
Target.

6)

The data flow for products should now look like this:

Figure 57:

Data flow for Products

PowerCenter 8.x Level I Developer

Lab 5-1

59

7)

Arrange All Iconic. The entire Mapping should look like this:

Figure 58:

8)

Completed Mapping m5_Load_ODS_PERSONNEL_STOCK_UNITS

Save your work and verify that the Mapping is valid.

Step 6.

Create and Run a Workflow

1)

In Workflow Manager, create a new Workflow named

wkf_Load_ODS_PERSONNEL_STOCK_xx.

2)

Add a Session task using the mapping you just completed and link it to the Start
task.

3)

Edit the Session. In the Mapping tab:

a)

Set the Relational Connection for the source SQ_EMPLOYEE_DEALERSHIP

to STGxx.

b)

Set the Relational Connection for the source SQ_SC_STG_PRODUCT to

STGxx.

c)

For SQ_SC_Inventory:
(i)

The value of Source file directory should be $PMSourceFileDir

(ii) The value of Source filename should be inventory.dat

60

Lab 5-1

PowerCenter 8.x Level I Developer

d)

For the targets SC_ODS_PERSONNEL and SC_ODS_STOCK_UNITS:

(i)

Verify that the Connection is ODSxx.

(ii) Set the Target load type to Normal.

(iii) Check the Truncate target table option.
e)

Click OK.

4)

Save your work.

5)

Start the Workflow.

6)

The Task Details and Source/Target Statistics for the completed Workflow should
look like this:

Figure 59:

Step 7.
1)

Statistics for wkf_Load_ODS_PERSONNEL_STOCK_xx

Preview Data to Verify Results

Preview the data for SC_ODS_STOCK_UNITS. It should look like this:

Figure 60:

Preview of ODS_STOCK_UNITS Table

PowerCenter 8.x Level I Developer

Lab 5-1

61

2)

Preview the data for ODS_PERSONNEL. It should look like this:

Figure 61:

62

Preview of ODS_PERSONNEL Table

Lab 5-1

PowerCenter 8.x Level I Developer

Extra Credit Lab:

Note that there are entries in ODS_PERSONNEL with null Employee IDs. Modify the part
of the Mapping that populates this table to substitute the default value 99999.
Hint: Use the Default Value property in the Expression transformation.

PowerCenter 8.x Level I Developer

Lab 5-1

63

64

Lab 5-1

PowerCenter 8.x Level I Developer

Lab 6-1: Using the Lookup Transformation

Scenario:
The ODS database needs a table that calculates such sales values as gross and net profit.
The sales executives want to know if more cars are sold on some weekdays than others.
Neither the sales date nor the name of the day of the week is in the Transaction table.
They must be added to the Sales table in the ODS.
Calculating gross and net profit is simplified by calculating an intermediate metric,
Margin.
Goals:

Use a Lookup transformation to import dates

Use a variable to calculate net and gross profit

Duration:
40 minutes

Instructions
Step 8.
1)

Create and Configure a Lookup Transformation

In PowerCenter Designer, create a new Mapping named
m6_Load_ODS_SALES_xx.
a)

2)

Drag in the source SC_STG_TRANSACTIONS and the target

SC_ODS_SALES

Click the Lookup transformation icon (

Designer workspace.
a)

Figure 62:

), then click again in the Mapping

Click the Import button and select From Relational Table

Import from Relational Table

PowerCenter 8.x Level I Developer

Lab 6-1

65

b)
Figure 63:

Import STG_DATES

3)

Drag the port DATE_ID from the Source Qualifier and drop it on the Lookup
transformation to create a link.

4)

Double-click the Lookup transformation to edit it.

a)

Rename the transformation lkp_STG_DATES_WEEKDAY.

b)

Click the Ports tab.

c)

Rename the port DATE_ID1 to DATE_ID_TRANS (to indicate it comes

from the transactions table).

d)

Click the Condition tab.

e)
Figure 64:

f)

66

In the Import Tables dialogue, log into the STGxx schema and select the
STG_DATES table.

Create a new condition. It should appear as shown.

Condition for lkp_STG_DATES_WEEKDAY

Click OK to close the Edit Transformations dialogue.

Lab 6-1

PowerCenter 8.x Level I Developer

5)

Connect the ports DATE_DESC and WEEKDAY_DESC to the Target ports

SALE_DATE and WEEKDAY_DESC respectively. The Mapping should now look
like this:

Figure 65:

Mapping with Lookup Configured

6)

Iconize the Lookup transformation and drag it out of the way.

7)

Save your work.

Step 9.

Create an Expression using a Variable Port

1)

Create an Expression transformation named exp_CALCULATE_FINANCIALS.

2)

Drag the following ports from the Source Qualifier to the Expression
transformation:
REVENUE
COST
DELIVERY_CHARGES
SALES_QTY
DISCOUNT
HOLDBACK
REBATE
a)

Figure 66:

The mapping should now look like this:

Mapping with Expression Added

PowerCenter 8.x Level I Developer

Lab 6-1

67

3)

Edit exp_CALCULATE_FINANCIALS and select the Ports tab.

a)

Create a new port called v_MARGIN, with Datatype decimal 10.2 and port
type Variable.

Note: the "v" indicates that this port is a variable.

(i)
Figure 67:

It should look like this:

Creating v_MARGIN

(ii) Open the Expression Editor for v_MARGIN.

(iii) Margin is calculated by multiplying SALES_QTY by the difference of
REVENUE and COST, as shown.

68

Lab 6-1

PowerCenter 8.x Level I Developer

Figure 68:

Formula for v_MARGIN

PowerCenter 8.x Level I Developer

Lab 6-1

69

b)

Create another new port named GROSS_PROFIT with Datatype decimal 10.2
and set it to Output only.
(i)

c)

Create another new port named NET_PROFIT with Datatype decimal 10.2
and set it to Output only.
(i)

d)
Figure 69:

Using the Expression Editor, set its formula to v_MARGIN(DELIVERY_CHARGES + DISCOUNT + HOLDBACK + REBATE)

The Ports tab should look like this:

Ports for exp_CALCULATE_FINANCIALS

e)

Click OK

f)

Save your work.

Step 10.

70

Using the Expression Editor, set its formula to v_MARGIN DELIVERY_CHARGES.

Finish the Mapping

1)

Connect the ports GROSS_PROFIT and NET_PROFIT from the Expression

transformation to the ports with the same names in the Target definition.

2)

Autolink by name from the Source Qualifier to the Target definition.

3)

Save your work and ensure that the Mapping is valid.

Lab 6-1

PowerCenter 8.x Level I Developer

Figure 70:

Step 11.

M6_Load_ODS_SALES

Create and Run the Workflow

1)

In Workflow Manager, create a Workflow called wkf_Load_ODS_SALES_xx.

2)

Add a Session task to the Workflow based on the Mapping

m6_Load_ODS_SALES_xx.

3)

Link the Start task to the Session task

4)

Edit the Session task. In the Mapping tab:

a)

The Source table needs the STGxx relational connection

b)

The target table needs the ODSxx relational connection.

(i)

Set the load type to normal

(ii) Select the Truncate table option

c)

Click the object lkp_STG_DATES_WEEKDAY

(i)

The Type of the connection should already be set to Relational.

(a) If not, set it.

(ii) Set the value of the Relational connection to STGxx.

5)

Click OK.

6)

Save your work.

7)

Run the workflow.

Figure 71:

Task Details and Source/Target Statistics for wkf_Load_ODS_SALES

PowerCenter 8.x Level I Developer

Lab 6-1

71

8)

Preview the data for the ODS_SALES table.

Figure 72:

72

Preview Results for ODS_SALES Table

Lab 6-1

PowerCenter 8.x Level I Developer

Lab 7-1: Using Aggregators and Sorters

Scenario:
In addition to the ODS_Sales table, sales executives need tables that summarize sales at
the weekly and monthly level. This will facilitate report generation and ad-hoc analysis.
Goals:

Use a Lookup transformation to add week and month information to the data rows
Split the data stream to feed two Aggregator transformations and write data out to two
separate tables
Use Sorter transformations to improve efficiency of the mapping
Duration:
75 minutes

Instructions
Step 1.

Create the Mapping

1)

Create a new Mapping named m7_Sales_Summaries_xx.

2)

Drag in the Source definition SC_ODS_SALES and the Target definitions

SC_ODS_SALES_BY_WEEK and SC_ODS_SALES_BY_MONTH.

3)

Add a Lookup transformation to the Mapping.

Note: If you need help with this step, consult the instructions for Lab 6-1.
a)
b)
c)

Import the table STG_DATES as its source.

Drag the port SALE_DATE from the Source Qualifier to the Lookup
transformation.
Edit the Lookup transformation.
(i)

Change its name to lkp_STG_DATES_WEEK_MONTH.

(ii) Set its Lookup condition to DATE_DESC = SALE_DATE

Step 2.
Create and Configure an Aggregator to Summarize Data by
Month
1)

Use the Aggregator icon (

Mapping.

PowerCenter 8.x Level I Developer

) to add an Aggregator transformation to the

Lab 7-1

73

2)

Drag the following ports to the Aggregator:

a)

From lkp_STG_DATES_WEEK_MONTH:
MONTH_DESC

b)

From the Source Qualifier SQ_SC_ODS_SALES:

REVENUE
HOLDBACK
COST
REBATE
DELIVERY_CHARGE
GROSS_PROFIT
SALES_QTY
NET_PROFIT
DISCOUNT

c)

From the Target SC_ODS_SALES_BY_MONTH:

REVENUE
HOLDBACK
COST
REBATE
DELIVERY_CHARGE
GROSS_PROFIT
SALES_QTY
NET_PROFIT
DISCOUNT

Note: PowerCenter will automatically rename these ports with a numerical 1 at the end of
each port name to avoid having duplicate port names.
3)

Edit the Aggregator.

a)

Rename it agg_SUM_BY_MONTH.

b)

Select the Ports tab.

c)
d)

Check the "Group By" checkbox for the port MONTH_DESC

For each of the ports imported from SC_ODS_SALES_BY_MONTH:
(i)

Change the numerical 1 in the name to _MONTH_SUM (so that, for

example, REVENUE1 becomes REVENUE_MONTH_SUM and so on).

(ii) Change the port type from input only to output only.
(iii) Edit each of the expressions for the _MONTH_SUM ports to calculate a
sum. For example, the expression for REVENUE_MONTH_SUM should
be SUM(REVENUE).

74

Lab 7-1

PowerCenter 8.x Level I Developer

e)
Figure 73:

The Aggregator should now look like this:

agg_SUM_BY_MONTH

4)

Click OK to exit the Edit Transformations dialogue.

5)

Link the MONTH_DESC port and the MONTH_SUM ports to the appropriate
ports in the Target SC_ODS_SALES_BY_MONTH.

6)

Save the Mapping. It may not validate at this time.

PowerCenter 8.x Level I Developer

Lab 7-1

75

7)

The Mapping should look like this:

Figure 74:

Mapping with One Aggregator Added

Step 3.
Create and Configure an Aggregator to Summarize Data by
Week
1)

2)

3)

Copy the Aggregator:

a)

Right-click the Aggregator and select "Copy"

b)

Right-click in the workspace and select "Paste"

Drag the following ports to the new Aggregator:

a)

Connect WEEK_DESC from lkp_STG_SALES_WEEK_MONTH to the

MONTH_DESC port of the new Aggregator

b)

Connect the following ports from the Source Qualifier SQ_SC_ODS_SALES

to the new Aggregator:
REVENUE
HOLDBACK
COST
REBATE
DELIVERY_CHARGE
GROSS_PROFIT
SALES_QTY
NET_PROFIT
DISCOUNT

Edit the new Aggregator.

a)

Rename it agg_SUM_BY_WEEK.

b)

Select the Ports tab.

c)
d)

Change the name of the port MONTH_DESC to WEEK_DESC

For each of the MONTH_SUM ports:
(i)

76

Change the name to _WEEK_SUM (so that, for example,

REVENUE_MONTH_SUM becomes REVENUE_WEEK_SUM and so
on)

4)

Click OK to exit the Edit Transformations dialogue.

5)

Connect the WEEK_DESC and WEEK_SUM ports of the new Aggregator to their
counterparts in the target SC_ODS_SALES_BY_WEEK.

Lab 7-1

PowerCenter 8.x Level I Developer

6)

Save the Mapping and make sure that it validates.

7)

The Mapping should look like this when you Arrange All Iconic:

Figure 75:

Step 4.

M7_Sales_Summaries

Create and Run the Workflow

1)

Create a new Workflow named wkf_Load_Sales_Summaries_xx

2)

The Session should both read from and write to Relational connection ODSxx.

3)

Set the relational connection for the Lookup to STGxx.

4)

Be sure to set the Load type to Normal and the Truncate table option on (checked).

5)

Save and run the Workflow.

Figure 76:

a)

Task Details and Source/Target Statistics for s_m9_Sales_Summaries

Note how long the Session task took to run:

PowerCenter 8.x Level I Developer

Lab 7-1

77

Figure 77:

Preview of Data in Table ODS_SALES_BY_MONTH

Figure 78:

Preview of Data in Table ODS_SALES_BY_WEEK

Note: The basic functionality of the Mapping is complete. However, in the Production
environment, where there will be millions of records, the Aggregator transformations
may run very slowly.
By default, Aggregator transformations work by creating a "bucket" for each unique value in
the Group By port(s). If the number of unique values is large, a great deal of memory
may be dedicated to maintaining these "buckets," or the system may have to cache
buckets to disk. In either case this can have a performance impact.
To prevent this, you can sort the data prior to its reaching the Aggregator. If the data is
sorted on the Group By port, and the Aggregator transformation is "told" that this is the
case, then there is no need to maintain many "buckets," and performance is improved.

Step 5.
1)

78

Add Sorters to the Mapping

Use the Sorter icon (

) to add a Sorter transformation to the mapping.

Lab 7-1

PowerCenter 8.x Level I Developer

2)

3)

Drag the following ports into it:

a)

From lkp_STG_SALES_WEEK_MONTH:
MONTH_DESC

b)

From SQ_SC_ODS_SALES:
REVENUE
COST
DELIVERY_CHARGE
SALES_QTY
DISCOUNT

HOLDBACK
REBATE
GROSS_PROFIT
NET_PROFIT

Edit the Sorter transformation.

a)

Rename it srt_MONTH_DESC

b)

Select the Ports tab.

(i)

Set the port MONTH_DESC as the Key to sort on.

(ii) Accept the default "Ascending" sort.

c)

The Sorter should now look like this:

Figure 79:

d)

Click OK to close the Edit Transformations dialogue.

4)

Remove all links from SQ_SC_ODS_SALES and

lkp_STG_DATES_WEEK_MONTH to the agg_SUM_BY_MONTH.

5)

Link the ports from srt_MONTH_DESC to agg_SUM_BY_MONTH.

6)

Edit agg_SUM_BY_MONTH.

PowerCenter 8.x Level I Developer

Lab 7-1

79

a)

Select the Properties tab.

b)

Check the box for "Sorted Input"

c)
7)

8)

Following the same procedure, create a Sorter transformation called

srt_WEEK_DESC and use it to sort the input to agg_SUM_BY_WEEK.
a)

Do not forget to set the Sorted Input box for agg_SUM_BY_WEEK.

b)

Alternatively, you can copy and edit the Sorter you have already created

The completed Mapping, when you Arrange All Iconic, should look like this:

Figure 80:

9)

M9_Sales_Summaries with Sorter Transformations

Save your work!

Step 6.
1)

Click OK.

Edit the Workflow

In Workflow Manager, right-click the session s_m7_Sales_Summaries and select
"Refresh Mapping."

Note: "Refresh Mapping" re-reads the mapping information for the Session. If substantial
changes have been made to the mapping that might cause it to become invalid, the
Workflow Manager marks it invalid just in case.

80

2)

Save the Workflow and verify that it is valid.

3)

Run the Workflow.

a)

You should get the same Source/Target statistics as before.

b)

But compare the run time for the Session to the first run.

Lab 7-1

PowerCenter 8.x Level I Developer

Lab 8-1: Using the Debug Wizard

Scenario:
The Mapping m8_STG_DATES_DEBUG contains at least one error that results in bad
data loaded into the target table. You must find and correct this error so that the data
warehouse project can proceed successfully.
Goals:

Use the Debug toolbar

Use the Debug Wizard

Duration:
35 minutes

Instructions
Step 1.

Copy and Inspect the Debug Mapping

1)

In the Designer application, make sure you are connected and open to your assigned
Devxx folder.

2)

Expand the folder SC_DATA_STRUCTURES

3)

a)

Locate and select the Mapping m8_STG_DATES_DEBUG.

b)

From the menu, select EditCopy.

Return to your Devxx folder.

a)

Select your Mapping subfolder.

b)

From the menu, select EditPaste.

c)
4)

In the "Copy Confirmation" dialogue, click Yes.

Save the Repository.

Tip: Note that the Mapping validates properly. The validation process ensures that the
Mapping is technically valid, but it cannot test for errors in business logic.
5)

6)

Open the Mapping in the Mapping Designer workspace.

a)

Inspect the Mapping to get an overall idea of what kind of processing is being
done.

b)

Read the expressions in the Expression transformation.

You have been told only that there is an "error" in the data being written to the
target, without any further clarification as to the nature of the error.

Tip: Many Mapping errors can be found by carefully inspecting the Mapping, without using
the Debug Wizard. If the error cannot be quickly located in this manner, the Debug
Wizard can help you by showing the actual data passing through the transformation
ports. However, to use the Debug Wizard effectively, you need to understand the logic
of the Mapping.

Step 2.

Open the Debug Toolbar and Start the Debug Wizard

PowerCenter 8.x Level I Developer

Lab 8-1

81

1)

On the Toolbar, right-click to bring up a list of available toolbars. Select the

"Debugger" toolbar.

Figure 81:

Open the Debugger Toolbar

Figure 82:

Debugger Toolbar

Tip: If the Debugger Toolbar is not visible, it is possible that another toolbar has shifted it
off the screen. Rearrange the other toolbars until you can see it.
2)

EITHER click the Start Debugger button (

Wizard.

) OR press F9 to start the Debug

3)

The first page of the Debug Wizard is informational. Please read it and press Next.

Tip: The Debug Wizard requires a valid Mapping and Session to run - it cannot help you
determine why a Mapping is invalid. The Output window of the Designer will show you
the reason(s) why a Mapping is invalid.

82

Lab 8-1

PowerCenter 8.x Level I Developer

4)

The Wizard should now look like this:

Figure 83:

a)

From the dropdown box, select the Integration Service

INT_SVC_EDW_DEV.

b)

In the Session box, select the Create a debug session radio button.

c)
5)

Debug Session Creation Dialogue

Click Next.

The next page of the Wizard allows you to set connection properties, similar to
creating Sessions in the Workflow Manager application.
a)

Set the Target Connection Value to STGxx

(i)

b)

You will discard the debugger data in a later step, so this value will be
ignored.

Select the Properties tab at the bottom.

(i)

Add the ".txt" extension to the Source filename property value.

(ii) Set the Target load type property to Normal

PowerCenter 8.x Level I Developer

Lab 8-1

83

c)
Figure 84:

d)

Your Wizard should now look like this:

Debug Session Connections Dialogue

Click Next three times.

These panels enable you to set which transformations in the Mapping you wish to monitor
in this debugging session, and set Session configuration information, such as a parameter
file or which connections the variables $Source and $Target correspond to.

84

Lab 8-1

PowerCenter 8.x Level I Developer

e)
Figure 85:

f)
6)

Ensure that Discard target data is checked.

Discard Target Data

Click Finish.

PowerCenter creates and initializes a Debug Session.

a)

The Debug Wizard automatically opens windows to view target and

transformation data. No data will be read until you are ready to view it.

Resize the Debugger Target Data Display and Debugger Data Display windows as needed.
A good guideline is to have them look something like this:
Figure 86:

Debugger Windows

PowerCenter 8.x Level I Developer

Lab 8-1

85

Step 3.
1)

Locate the Error

Your screen should now look like this:

Figure 87:

Designer During a Debug Session

2)

In the Target Instance dropdown box, select STG_DATES_VIEW.

3)

In the Instance dropdown box, select SQ_dates.

Note: The term "instance" here refers to an object in the Mapping. Thus, each
transformation is an "instance."
4)

Click the "Next Instance" button (

a)

Figure 88:

5)

Data in the Instance Window

Note that there is no data available as yet - the Instance window, with the Next
Instance button, shows data as it moves from transformation to transformation
through the Mapping.

Click the Step to Instance button (

a)

86

Data displays in the Instance window.

Toggle the dropdown to the Expression transformation.

a)

6)

) on the Debugger toolbar.

Note that one more row has been read, and the first row has been "pushed" to
the Expression transformation and the Target table.

Lab 8-1

PowerCenter 8.x Level I Developer

7)

Click the Step to Instance button several more times (at least 13), watching how the
data flows from the Instance window to the Target Instance window. Compare the
results between the Target instance and Instance windows.
a)

What is the nature of the error in the data being written to the table?

b)

Double-click the Expression transformation to open it.

Tip: Note that the transformation properties are grayed-out. While you can view and copy
expressions, you cannot edit the Mapping or its components while the Debugger session
is running.
c)

Step 4.

What is causing the error?

Fix the Error

Tip: Nonetheless, you CAN try new variations on expression while the Debugger is running.
1)

Click the Ports tab.

a)

Enter the Expression Editor for one of the output ports - preferably the one
that seems likely to be causing the problem.

b)

Select the text of the expression (even though it is grayed-out) and copy it to the
Windows clipboard by typing Ctrl+C.

c)

Click Cancel twice.

d)

Right-click the Expression transformation and select Evaluate Expression.

e)

Paste the expression text you chose into the Expression Editor and press
Evaluate.
(i)

The Debugger will immediately evaluate the expression with the current
data in the ports.

(ii) You can make as many changes to the Expression here as you need.
(iii) Once you have a modified expression that you want to keep, copy it to the
Windows clipboard.
2)

Close the expression evaluator

3)

Stop the Debugger by pressing the Stop Debugger button (

toolbar
a)

) on the Debugger

Click Yes to "Shutdown the debugger."

PowerCenter 8.x Level I Developer

Lab 8-1

87

88

4)

Edit the Expression transformation and put your modified Expression in place by
pasting it into the Expression Editor.

5)

Save your work.

6)

Restart the Debugger and test to ensure that your fix worked.

Lab 8-1

PowerCenter 8.x Level I Developer

PowerCenter 8.x Level I Developer

Lab 8-1

89

Answers
3.7.a What is the nature of the error in the data being written to the table?
The month and date seem to be reversed. That is, the data comes in as January 1, January 2, etc., but is
being written as January 1, February 1, etc.
3.7.c. What is causing the error?
The Expression Editor is using a format of DD/MM/YYYY but the incoming data has a format of
MM/DD/YYYY.

90

Lab 8-1

PowerCenter 8.x Level I Developer

Lab 9-1: Updating Target Tables

Scenario:
The ODS_Customer table needs to be updated periodically with a list of customers that
have recently done business with Mersche Motors.
Many customers will be duplicates of existing customers. Some will be the same people
but with new information.
You need to devise a Mapping that will identify the new customers and add them to the
ODS table.
Also, if a customer's information has been updated, the ODS table needs to be updated
as well.
Goals:

Use a single Source definition to read two files and combine their data in a single data
stream.
Remove duplicate rows.
Create logic that
o
Rejects the record if the incoming CUSTOMER_ID is missing
o
Inserts the record if the customer does not already exist in ODS_CUSTOMERS
o
Updates the record if the customer already exists in ODS_CUSTOMERS
Duration:
90 minutes

Instructions
Step 1.

Sources and Targets

1)

Create a new Mapping named m9_Update_Customers_xx.

2)

Drag the Source definition SC_Customers_updated into the mapping.

3)

4)

a)

Rename the Source definition as SC_Customers_updated_East

b)

Rename the Source Qualifier as SQ_SC_Customers_updated_East

Drag the Source definition SC_Customers_updated into the mapping again

a)

Rename the Source definition as SC_Customers_updated_West

b)

Rename the Source Qualifier as SQ_SC_Customers_updated_West

Arrange All

PowerCenter 8.x Level I Developer

Lab 9-1

91

5)

The Mapping should look like this:

Figure 89:

Step 2.

Two Sources Based on One Source Definition

Merge the Two Data Streams Using a Union Transformation

1)

Use the Union icon (

2)

Edit the Union transformation.

a)

Rename it uni_E_W_CUSTOMERS.

b)

Select the Groups tab.

(i)

Figure 90:

c)
3)

92

) to add a Union transformation to the Mapping.

Configure two groups named EAST_CUSTOMERS and

WEST_CUSTOMERS.
Union Groups

Click OK.

Select all the ports in SQ_SC_Customers_updated_East and drag them to the

Union transformation.

Lab 9-1

PowerCenter 8.x Level I Developer

a)
Figure 91:

4)

The Mapping should now look like this:

Mapping with Union

Link the ports from SQ_SC_Customers_updated_West to the

WEST_CUSTOMERS group of the Union transformation.

Step 3.

Remove Duplicates with a Sorter Transformation

1)

Add a Sorter transformation to the Mapping.

2)

Drag all the output ports from the Union transformation to it.

3)

Edit the Sorter transformation.

a)

Rename it srt_REMOVE_CUST_DUPS.

PowerCenter 8.x Level I Developer

Lab 9-1

93

b)
Figure 92:

c)
d)
Figure 93:

e)

94

Sort on all the ports.

Sorting on All Ports

Select the Properties tab.

Set the Distinct checkbox to true (checked)
Distinct Checkbox Set to True

Click OK.

Lab 9-1

PowerCenter 8.x Level I Developer

f)
Figure 94:

4)

The Mapping should look like this:

Mapping with Sorter

Iconize the Source definitions, Source Qualifiers, and Union transformation, and
arrange the Mapping to give you space on the right.

Step 4.
Create and Configure a Lookup on the ODS_CUSTOMERS
target table
1)

Drag the Target SC_ODS_CUSTOMERS into the Mapping.

2)

Add a Lookup transformation to the Mapping.

a)

3)

Use the relational table SC_ODS_CUSTOMERS.

Drag the port CUSTOMER_NO from the Sorter transformation to the Lookup
transformation.

Note: the rule of "active vs. passive" transformation objects applies here. The Sorter is an
active transformation. Therefore, it cannot be bypassed by bringing this port directly
through from the Union transformation to the Target.
4)

Edit the Lookup transformation.

a)

Rename it lkp_ODS_CUSTOMERS.

b)

Rename the port CUSTOMER_NO1 to CUSTOMER_NO_SOURCE

c)

Set the Lookup condition to

CUSTOMER_NO = CUSTOMER_NO_SOURCE

d)

Click OK.

PowerCenter 8.x Level I Developer

Lab 9-1

95

5)

The Mapping should now look like this:

Figure 95:

Step 5.

96

Mapping with Lookup

Create and Configure an Update Strategy Transformation

1)

Use the Update Strategy icon (

Mapping.

) to add an Update transformation to the

2)

Drag all the ports from the Sorter transformation to the Update Strategy
transformation.

3)

Drag the CUSTOMER_NO port from the Lookup transformation to the Update
Strategy transformation.

4)

Edit the Update Strategy transformation.

a)

Rename it to upd_UPDATE_ELSE_INSERT.

b)

Select the Ports tab.

c)

Change the name of the CUSTOMER_NO port (the one coming from the
Sorter) to CUSTOMER_NO_SOURCE

d)

Change the name of the CUSTOMER_NO1 port (the one coming from the
Lookup) to CUSTOMER_NO_LOOKUP

Lab 9-1

PowerCenter 8.x Level I Developer

Figure 96:

Update Strategy Transformation with Port Names Changed

PowerCenter 8.x Level I Developer

Lab 9-1

97

e)

Select the Properties tab.

f)

In the Update Strategy Expression field, enter the Expression Editor.

Update Strategy Expression

The goal is to
Reject the row if the customer number is null
Insert the row if the customer number is new
Update the record if the customer number already exists.
Starting with the case where the Customer number coming from the Source is null, we want
to reject the record:
IIF(ISNULL(CUSTOMER_NO_SOURCE), DD_REJECT
Then we proceed to the case where the customer number from the source exists and does not
exist in ODS_CUSTOMER, meaning it is a new customer. Then we want to insert a new
record:
IIF(ISNULL(CUSTOMER_NO_LOOKUP), DD_INSERT
If neither of these is true, then the customer number already exists in ODS_CUSTOMER,
so we want to update it:
DD_UPDATE
When we put it all together, the expression becomes:
IIF(ISNULL(CUSTOMER_NO_SOURCE), DD_REJECT,
IIF(ISNULL(CUSTOMER_NO_LOOKUP), DD_INSERT, DD_UPDATE))
g)

Enter the above expression in the Expression Editor and click Validate.
(i)

Figure 97:

h)
5)

98

If the expression does not validate, fix it.

Properties of the Update Strategy Transformation

Click OK to exit the Edit Transformations dialogue.

Use Autolink by Position to connect the ports from the Update Strategy
transformation to the Target definition.

Lab 9-1

PowerCenter 8.x Level I Developer

6)

Arrange All Iconic to view the logic of the Mapping.

Figure 98:

Step 6.

M9_Update_Customers_xx

Create and Run the Workflow

1)

Create a Workflow named wkf_Update_ODS_CUSTOMERS_xx

2)

The source files will be customers_updated_east.dat and

customers_updated_west.dat

3)

The Target and Lookup transformation will use the Relational connection ODSxx

4)

Make sure to set Target load type to Normal.

5) DO NOT set the Truncate target table option. If you do, the existing data in the
table will be deleted and the update logic will not work properly.
Figure 99:

Source/Target Statistics for wkf_Update_ODS_Customers_xx

Note 1: The specific number of rows may vary depending on whether you did the Extra
Credit exercise in Lab 3-1.
Note 2: The number of rejected rows shown here does not reflect the number of rows
rejected by the Mapping. Rather, it shows that no errors were thrown by the database.
This is to be expected because the Mapping did not forward any rows with a null key
field to the database. To see the number of rows actually rejected by the Mapping, you
must consult the Session Log.

PowerCenter 8.x Level I Developer

Lab 9-1

99

100

Lab 9-1

PowerCenter 8.x Level I Developer

Lab 10-1: Filling in Missing Data

Scenario:
IT and test users have noted that there are invalid and missing Postal codes for some
customers. This hampers Marketing's efforts to identify who Mersche Motors customers
are and how to reach out to them, and must be corrected.
Marketing wishes customers to be categorized based on a combination of income level
and location.
Marketing wants a file of customers with subprime credit ratings that can be analyzed
further.
Goals:

Use an Unconnected Lookup transformation to attempt to find in which city customers

purchased their cars.
o
This works on the assumption that customers purchase their cars locally.
Use a Router transformation to categorize customers on income level, Postal code, and
city into High Income, Sub-Prime, and Uncategorized categories.
o
Since the reality is that Sub-Prime and Uncategorized will be treated the same from
a lending standpoint, write them to the same table.
Override the Sub-Prime/Uncategorized relational writer to write out to a flat file (.CSV)
so that the data can easily be loaded into a spreadsheet for further analysis.

Duration:
60 minutes

Instructions
Step 1.

Create a Mapping, Select Sources and Targets

1)

Create a Mapping called m10_Customer_Category_xx.

2)

Add the Source definition SC_ODS_CUSTOMERS.

3)

Add the Target definitions SC_ODS_CUSTOMERS_HIGH_VALUE and

SC_ODS_CUSTOMERS_SUBPRIME.

4)

Save the Mapping (it will not be valid).

PowerCenter 8.x Level I Developer

Lab 10-1

101

Step 2.
1)

Identify Bad Data and Establish a Method of Remediation

Preview the data in ODS_CUSTOMERS.

Figure 100: Postal Code Data

2)

Note that there are bad entries in the Postal Code field, such as:
0.000
2112.
NULL (string)

3)

Not every record will require an attempt to repair bad Postal Code.
a)

4)

Close the Preview Data window.

Step 3.

Create and Configure a Lookup Transformation

1)

Create a Lookup transformation named lkp_RETURN_LOCATION.

2)

Base the Lookup on the SC_STG_DEALERSHIP table.

3)

Drag DEALERSHIP_ID from the Source Qualifier to the Lookup.

4)

Edit the Lookup transformation.

a)

102

Therefore, we will use an Unconnected Lookup as needed to fix the data.

Rename the Port from DEALERSHIP_ID1 to DEALERSHIP_ID_SOURCE

b)

Set the Lookup condition as

DEALERSHIP_ID = DEALERSHIP_ID_SOURCE

c)

Set DEALERSHIP_LOCATION as the Return port.

Lab 10-1

PowerCenter 8.x Level I Developer

d)

Click OK.

5)

Delete the link from the Source Qualifier to the Lookup transformation.

6)

The Mapping should now look like this:

Figure 101: Mapping with Unconnected Lookup Added

Step 4.

Create and Configure an Expression Transformation

1)

Create an Expression transformation.

2)

Drag all the ports from the Source Qualifier into the Expression transformation.

3)

Edit the Expression transformation.

a)
b)
c)

Rename it exp_FIND_LOCATION.
Create a new Output port named LOCATION with datatype STRING and
precision 20.
Enter the Expression Editor for the port LOCATION.
(i)

In the Functions tab, scroll to the bottom of the list of functions.

(ii) At the bottom of the list you will see a folder named "Lookups." Open this
folder and you will see the unconnected Lookup transformation you just
created.
(iii) Create the expression so that if the value of POSTAL_CODE is NULL, or
contains the string NULL, or is equal to 0.000, then look up the
POSTAL_CODE in the STG_DEALERSHIP table based on
DEALERSHIP_ID. Otherwise, return the value of POSTAL_CODE.
IIF(((POSTAL_CODE != '0.000') AND NOT
ISNULL(POSTAL_CODE) AND (POSTAL_CODE != 'NULL')),
POSTAL_CODE,
:LKP.LKP_RETURN_LOCATION(DEALERSHIP_ID) )
(iv) Make sure your expression is valid, then click OK.
d)

Click OK to exit the Edit Transformations dialogue.

PowerCenter 8.x Level I Developer

Lab 10-1

103

Step 5.
Create and Configure a Router Transformation to Classify
Customers
)

1)

Create a Router transformation (

2)

Drag all ports from the Expression transformation to the Router transformation
except DEALERSHIP_ID.

3)

Edit the Router transformation.

a)

Rename it rtr_CLASSIFY_CUSTOMERS.

b)

Create two Groups, named HIGH_VALUE and SUBPRIME.

c)

For the HIGH_VALUE group set the filter condition to send records to this
group when
the High Income flag is set to 1
OR
the value in the Location port is 19104, 10005, 90004, Newport Beach,
Scottsdale, or West Palm Beach

d)

For the SUBPRIME group set the condition to send records to this group when
the High Income flag is set to 0
AND
the value in the Location port is 55409, 98112, 75201, Indianapolis, or
Phoenix.

Note: Obviously these location choices are very simplistic. They are acceptable for
illustrating the use of the Router transformation in this exercise, however.
e)

Step 6.

Verify that your expressions are valid.

Connect Router Output Groups to Targets

1)

Connect the appropriate HIGH_VALUE output group ports to the target

SC_ODS_CUSTOMERS_HIGH_VALUE

2)

Connect the SUBPRIME group to the target

SC_ODS_CUSTOMERS_SUBPRIME

3)

To connect the Default group (uncategorized records)

a)

Drag in another copy of SC_ODS_CUSTOMERS_SUBPRIME

b)

Rename it SC_ODS_CUSTOMERS_UNCATEGORIZED

Note: Even though the Target instance has been renamed, it will still write to the original
table name. You can verify this by looking at the Shortcut To fields.
c)
4)

104

Connect the DEFAULT group from the Router to the Target

SC_ODS_CUSTOMERS_UNCATEGORIZED

Save your work.

Lab 10-1

PowerCenter 8.x Level I Developer

5)

Arrange All Iconic.

Figure 102: M10_Customer_Category_xx

Step 7.
1)

Create and Run a Workflow for the Mapping

Name the Workflow wkf_Customer_Category_xx.

Note: You will override the Relational Writer so that the Subprime and Uncategorized
customers are written to .csv files.
2)

Edit the Session object.

a)

Select the Mapping tab.

b)

Select the Target SC_ODS_CUSTOMERS_SUBPRIME

c)
d)
e)

Using the drop box, set the value of the Writers property to File Writer.
In the Properties window, scroll down to find the Output Filename attribute.
Change its value to CUSTOMERS_SUBPRIME_xx.csv

PowerCenter 8.x Level I Developer

Lab 10-1

105

f)

The Edit Tasks window should look like this:

Figure 103: File Output for Subprime Customers

3)

In the "Properties" bar, click the link Set File Properties

a)
b)

In the "Flat Files - Targets" dialogue, select Delimited and click Advanced.
In the "Delimited File Properties - Targets" dialogue, make sure that the
Column Delimiter is a comma (,) character.

c)

Click OK

d)

Click OK.

4)

Using the same procedure, set the Target

SC_ODS_CUSTOMERS_UNCATEGORIZED to output to a file named
CUSTOMERS_UNCATEGORIZED_xx.csv

5)

Save and run the Workflow.

Figure 104: Source/Target Statistics

106

Lab 10-1

PowerCenter 8.x Level I Developer

6)

In Windows Explorer, navigate to C:\pmfiles\TgtFiles (or the equivalent using your

mapped drive, InfaShared on IntDevTest).

7)

Double-click the file CUSTOMERS_SUBPRIME_xx.csv to open it in MS Excel.

(If you don't have MS Excel, then use an editor of your choice such as Notepad.)
a)

It should look like this:

Figure 105: Subprime Customers Output File

Extra Credit
1. Extend the invalid POSTAL_CODE search to include fields that have a period character (.)
2. Redesign the mapping so that all of the POSTAL_CODE values are replaced with city names.

PowerCenter 8.x Level I Developer

Lab 10-1

107

108

Lab 10-1

PowerCenter 8.x Level I Developer

Lab 10-2: Using a Persistent Variable

Scenario:
New employees need to be added to ODS_EMPLOYEES
They must have unique IDs assigned to them
Goals:

Further leverage the reusable transformation EXP_Format_Records

Use a persisted Repository variable to hold and increment the latest employee
identification number

Duration:
45 minutes

Instructions
Step 1.

Create a Mapping Variable

1)

Create a new Mapping named m10_Load_New_Employees_Summary_xx.

2)

Drag in the flat file Source definition SC_Employees.

3)

Drag in the relational Target definition

SC_ODS_NEW_EMPLOYEE_SUMMARY.

4)

From the menu select Mappings Parameters and Variables.

a)

Create a new Variable with the following properties:

Name:
$$New_ID
Type:
Variable
Datatype:
Integer
Aggregation Type:
Count
IsExprVar:
FALSE
Initial value:
84699

Note: This variable will be incremented and used to generate new employee IDs.

PowerCenter 8.x Level I Developer

Lab 10-2

109

Figure 106: Declaring $$New_ID

Step 2.

Add and Edit the Expression Transformation

In this step, you will make a non-reusable copy of the reusable transformation you created
earlier in the course, and edit it.
1)

In your folder, expand the Transformations subfolder.

2)

Make a non-reusable copy of a reusable transformation

a)

Select the re_exp_Format_Persons transformation

b)

Click and hold the click, then hold down the CTRL key.

c)

Now drag the transformation into the open mapping,

m13_Load_New_Employees_Summary_xx.

d)

Notice that while you are doing this the Status Bar reads, "Make a non-reusable
copy of this transformation and add it to this mapping."

e)
3)

4)

This is how you make a non-reusable copy of a reusable transformation.

Edit the Expression transformation.

a)

Change its name to exp_Format_New_Employees

b)

Delete the ports Income and High Income Flag.

c)

Add a new output port named NEW_EMPLOYEE_ID, datatype Integer.

Edit the expression for the new port.

a)

Open the expression editor and delete the existing expression

(NEW_EMPLOYEE_ID).

b)

Select the Functions tab.

(i)

110

Scroll to the bottom of the Functions list.

Lab 10-2

PowerCenter 8.x Level I Developer

(ii) Open the folder named Variables.

(iii) Double-click the function SetCountVariable to place it in the Expression
Editor.
c)

Select the Variables tab.

(i)

Open the folder named Mapping Variables.

(ii) Double-click the variable $$New_ID to add it to the expression.

(iii) The Expression should now read:
SETCOUNTVARIABLE($$New_ID)
(iv) Click OK to validate the expression and exit the Expression Editor.
d)
5)

6)

7)

Click OK to exit the Edit Transformations dialogue.

Link the following ports from the Source Qualifier to the Expression
transformation:
From Port

To Port

FIRST_NAME

FIRSTNAME

LAST_NAME

LASTNAME

GENDER

GENDER

AGE

AGE

Link the following ports from the Expression transformation to the Target:
From Port

To Port

NAME

NAME

GENDER_CATEGORY

GENDER_CATEGORY

SENIOR_FLAG

SENIOR_FLAG

NEW_EMPLOYEE_ID

EMPLOYEE_ID

The Mapping should now look like this:

Figure 107: m10_Load_New_Employees_Summary_xx

PowerCenter 8.x Level I Developer

Lab 10-2

111

Step 3.

Create a Workflow and Run the Mapping

a)

Name the Workflow wkf_Load_New_Employees_Summary_xx.

b)

The source file is named employees_new.dat.

c)

The target relational connection is ODSxx.

d)

DO NOT set the Truncate option.

Figure 108: Source/Target Statistics for s_m10_Load_New_Employees_Summary

Step 4.
1)

Verify the Mapping Variable Incremented Properly

Preview the Target table.
a)

Set the Show up to field to 200, to ensure that all 109 rows are visible.

b)

Scroll to the bottom.

Figure 109: Preview Data for New Employees

c)
2)

112

The last employee ID showing should be 84808.

Change to the Workflow Manager

Lab 10-2

PowerCenter 8.x Level I Developer

3)

Right-click the Session object and select View Persistent Values.

Figure 110: Persistent Values for s_m10_Load_New_Employees_Summary_xx

a)

Step 5.

Note that the value of $$New_ID is the same as the value of the last
EMPLOYEE_ID. It is ready for the next run of the workflow.

Reconfigure and Rerun the Workflow

The purpose here is to verify that the counter is working properly.

1)

Change the Source file to employees_new2.dat

2)

Save and start the Workflow.

3)

Viewing the Source/Target statistics, note that the Source file contained 5 rows that
were added to the Target.

4)

View the Persistent Values for the Session and verify that the number has
incremented by five.

5)

Preview the data in the Target and verify that five new employees have been added
with the appropriate Employee ID numbers.

Extra Credit
If the Mapping had a relational source, how could a similar technique be used to read the Source
incrementally, so that only new records would be read each time the Session was run?

PowerCenter 8.x Level I Developer

Lab 10-2

113

Extra Credit Answer

If the Mapping had a relational source, how could a similar technique be used to read the Source
incrementally, so that only new records would be read each time the Session was run?
You can use a Variable to hold a date/time stamp, which you would put into a SQL override and filter all
records prior to that date/time.
A similar technique can be used to capture other ascending numbers such as DI, Oracle rownum, or SQL
Identity fields.

114

Lab 10-2

PowerCenter 8.x Level I Developer

Lab 11-1: Creating a Mapplet

Scenario:
The team lead has noticed that there are situations where some of the transformations
you developed in m8_Sales_Summaries_xx can be reused.
To take advantage of these previously created objects, you will create a Mapplet from
existing objects, which can then be used in other Mappings.
Goals:

Create a Mapplet

Duration:
40 minutes

Instructions
Step 1.

Copy and Prepare a Mapping

1)

Make a copy of the Mapping m7_Sales_Summaries_xx.

2)

Name it m11_Sales_Summaries_xx.

3)

Open the Mapping m11_Sales_Summaries_xx.

4)

Arrange All Iconic if the Mapping isn't already arranged that way

Figure 111: Mapping m11_Sales_Summaries

Step 2.

Create a Mapplet

1)

Select the following transformations:

lkp_STG_DATES_WEEK_MONTH
srt_MONTH_DESC
srt_WEEK_DESC
agg_SUM_BY_MONTH
agg_SUM_BY_WEEK

2)

From the menu, select EditCopy.

PowerCenter 8.x Level I Developer

Lab 11-1

115

3)

Open the Mapplet Designer tool (

).

4)

From the menu, select EditPaste.

5)

Name the new Mapplet mplt_Sales_Summaries.

6)

In the workspace, Arrange All.

Figure 112: New Mapplet

7)

Add a Mapplet Input transformation (

8)

Drag ports to the Mapplet Input transformation:

9)

).

a)

Drag the SALE_DATE port from the Lookup transformation to the Mapplet
Input transformation.

b)

Drag every input port except WEEK_DESC from srt_WEEK_DESC to the

Mapplet Input transformation.

c)

Drag every input port except MONTH_DESC from srt_MONTH_DESC to

the Mapplet Input transformation.

Edit the Mapplet Input transformation.

a)

Rename it in_Sales_Summaries.

b)

To ports REVENUE through NET_PROFIT, add the suffix _WEEK.

Hint: You can use your mouse and Ctrl+C (copy) and Ctrl+V (Paste) to speed the process.

116

c)

On ports REVENUE1 through NET_PROFIT1, replace the 1 with the suffix

_MONTH.

d)

Click OK to exit the Edit Transformations dialogue.

Lab 11-1

PowerCenter 8.x Level I Developer

Figure 113: Mapplet with Input Transformation Added

10) Add a Mapplet Output transformation (

a)
b)

) to the Mapplet.

Rename it out_Sales_Summary_Weekly.
From agg_SUM_BY_WEEK, drag ports WEEK_DESC and
REVENUE_WEEK_SUM through NET_PROFIT_WEEK_SUM to the
Mapplet Output transformation.

11) Add a second Mapplet Output transformation to the Mapplet.

a)
b)

Rename it out_Sales_Summary_Monthly.
From agg_SUM_BY_MONTH, drag ports MONTH_DESC and
REVENUE_MONTH_SUM through NET_PROFIT_MONTH_SUM to the
Mapplet Output transformation.

12) Save the Mapplet. Make sure it is valid.

Figure 114: Completed Mapplet

PowerCenter 8.x Level I Developer

Lab 11-1

117

Figure 115: Mapplet Arranged Iconic

Step 3.

Add the Mapplet to the Mapping

1)

Switch to the Mapping Designer, where m11_Sales_Summary should still be open.

2)

Delete the following transformations from the Mapping:

lkp_STG_DATES_WEEK_MONTH
srt_WEEK_DESC
srt_MONTH_DESC
agg_SUM_BY_WEEK
agg_SUM_BY_MONTH

3)

Drag the Mapplet mplt_Sales_Summaries into the mapping.

4)

Manually link from the port SALE_DATE in the Source Qualifier to the port
SALE_DATE in the in_SALES_SUMMARIES section of the Mapplet.

Hint: You may want to stretch the Mapplet vertically to see as many ports as possible.
5)

118

You will use Autolink to connect the remaining input ports in the Mapplet.
a)

Right-click in the workspace and select Autolink.

b)

In the From Transformation dropdown box, select the Source Qualifier.

c)

In the To Transformation box, if necessary expand the Mapplet, and select the
Input section in_Sales_Summaries.

d)

Click More.

e)

Enter _WEEK in the Suffix field under the To Transformation column, as

shown:

Lab 11-1

PowerCenter 8.x Level I Developer

Figure 116: Auto Link dialogue

f)

Click OK.

g)

Half the ports are linked.

h)

Repeat the process with the suffix _MONTH to complete the links to the Input
section of the Mapplet.

Figure 117: SQ Linked to Mapplet

6)

Now you will link the Mapplet to the Targets.

a)
b)

In the Out_Sales_Summary_Monthly section, click MONTH_DESC.

Hold down the SHIFT button on the keyboard and click
NET_PROFIT_MONTH_SUM to select all the fields in the section.

PowerCenter 8.x Level I Developer

Lab 11-1

119

c)

Drag the mouse from MONTH_DESC to MONTH_DESC in the Target

definition SC_ODS_SALES_BY_MONTH.

Tip: This can work because the ports are in exactly the same order on both the Mapplet
output section and the Target. It is equivalent to Autolink by Position, but does not
automatically start with the first port on each transformation.
d)
7)

Using the same procedure, link the ports of the Out_Sales_Summary_Weekly

to the Target SC_ODS_SALES_BY_WEEK.

Save the Mapping and make sure it is valid.

Figure 118: m11_Sales_Summaries_xx

120

Lab 11-1

PowerCenter 8.x Level I Developer

Lab 12-1: Using Assignment and Decision Tasks

Scenario:
Management wants the ability to report sales on a weekly basis.
You created a Mapping, (m6_Load_ODS_SALES_xx), that enables this reporting.
You need to put this Mapping into production with a more formal and fault-tolerant
Workflow.
Data must be retrieved from the OLTP source.
Goals:

Create a more formal Workflow that prevents some types of bad data from getting into
the ODS_SALES table.
Assign Workflow variables to keep track of the number of times the Workflow has been
run.
Increment Workflow variables using an Assignment task.
Branch in a Workflow using link conditions and a Decision task to choose to run the
next Session or report an error.
Duration:
45 minutes

Instructions
Step 1.
1)

Make a Session Reusable and Copy a Reusable Session

Make a Session Reusable
a)

In the Workflow Manager application, locate the Workflow in your folder

named wkf_Load_ODS_SALES_xx and drag it into the Workflow Designer
workspace.

b)

Edit the Session

(i)

In the General tab, check the "Make reusable" checkbox.

(ii) Click Yes to make the Session reusable.

(iii) Click OK.
c)
d)
2)

Save the Workflow.

Disconnect from your ~Devloperxx folder.

Use the R button (

) to switch to the Repository Manager application.

a)

Open your ~Developerxx folder.

b)

Open the SC_DATA_STRUCTURES folder.

c)

Using the EditCopy technique shown in an earlier lab, copy the reusable
Session s_m_Load_STG_TRANSACTIONS from
SC_DATA_STRUCTURES to your ~Developerxx folder.

d)

Disconnect from the repository. (The Repository Manager automatically saves

your folder.)

PowerCenter 8.x Level I Developer

Lab 12-1

121

3)

Switch back to the Workflow Manager application.

a)
b)

c)

Open your folder.

Verify that there are two reusable Sessions in the Sessions subfolder:
s_m_Load_STG_TRANSACTIONS and
s_m8_Load_ODS_SALES_xx.
Select the Task Developer tool
(i)

Drag the Session s_m_Load_STG_TRANSACTIONS into the workspace.

(ii) Edit the Session and change its name by adding the suffix _xx.

Step 2.

Create and Configure the Workflow

1)

Create a new Workflow named wkf_Load_ODS_SALES_from_OLTP_xx

2)

In the Create Workflow dialogue, select the Variables tab.

a)

Create a new Workflow variable with the following properties:

Variable Name =$$WORKFLOW_RUNS

Datatype = integer
Persistent = checked
Default value = 0

Figure 119: Create the Workflow Variable

122

3)

Drag the Session s_m_Load_STG_TRANSACTIONS_xx into the workflow and

link the Start task to it.

4)

Set the Relational reader for the source to SDBU.

5)

Set the Relational writer for the target to STGxx.

6)

Save your work.

Lab 12-1

PowerCenter 8.x Level I Developer

Step 3.

Create the Assignment Task

1)

Use the Assignment Task button (

2)

Link the Session task to the Assignment task.

3)

Double-click the link to edit it.

4)

Add a link condition to ensure that the Assignment task executes only if the Session
task was successful.
a)

) to add an Assignment task to the Workflow.

Select the pre-defined function "Status" and set the condition so that the status
must be SUCCEEDED. (See figure.)

Figure 120: Link Condition

5)

Edit the Assignment task.

a)
b)

Change its name to asgn_WORKFLOW_RUNS.

In the Expressions tab, create an expression that increments the User Defined
Variable by one. (See figure.)

Figure 121: Assignment Task Expression

6)

Save your work.

PowerCenter 8.x Level I Developer

Lab 12-1

123

Step 4.

Create the Decision Task

1)

Use the Decision task button (

2)

Link asgn_WORKFLOW_RUNS to the Decision task.

3)

Edit the link.

a)

4)

) to add a Decision task to the Workflow.

Add a condition to ensure that the Decision task executes only if

assgn_WORKFLOW_RUNS completed successfully.

Edit the Decision task

a)
b)

Rename it dcn_RUN_WEEKLY.
In the Properties tab, create a Decision Name expression to see if this is the
seventh day of the Workflow week.
(i)

The Modulus function (MOD) divides two numbers and yields the
remainder.

(ii) See the figure.

Figure 122: Decision Task Expression

Tip: The decision task evaluates an expression and returns a value of either TRUE or
FALSE. This value can be checked in a Link condition to determine the direction in
which the Workflow proceeds from the Decision task.
5)

Step 5.

124

Save your work.

Add a Second Session

1)

Drag the session s_m6_Load_ODS_SALES_xx into the Workflow.

2)

Link dcn_RUN_WEEKLY to it.

3)

Double-click the link.

Lab 12-1

PowerCenter 8.x Level I Developer

a)

Add a link condition that checks whether dcn_RUN_WEEKLY has returned a

value of TRUE. (See figure.)

Figure 123: Link Condition

4)

The Session properties were set correctly in the Workflow where you first created
this Session.

Step 6.

Create the Email Task

1)

Use the email task button (

2)

Link dcn_RUN_WEEKLY to the Email task.

3)

Add a Link condition that checks whether dcn_RUN_WEEKLY has returned a

value of FALSE.

4)

Edit the Email task.

a)
b)

) to add an Email task to the mapping.

Rename it eml_DAILY_LOAD_COMPLETE
In the Properties tab, enter appropriate values for Email User Name, Email
Subject, and Email Text (see example below).

Figure 124: Email Properties

c)

Click OK.

PowerCenter 8.x Level I Developer

Lab 12-1

125

d)

Right-click in the workspace and select Arrange Horizontal

Figure 125: wkf_Load_ODS_SALES_From_OLTP completed

e)

Step 7.

Save your work.

Start the Workflow and Monitor the Results

You will need to run the Workflow seven times in order to test the weekly aggregate session.
1)

Start the Workflow.

a)

Review the Workflow results in the Gantt view of the Workflow Monitor. It
should appear similar to the figure below:

Figure 126: Gantt Chart of the Completed Workflow Run

2)

126

Return to the Workflow Manager.

Lab 12-1

PowerCenter 8.x Level I Developer

3)

In the Navigator window, in the Workflows subfolder, right-click

wkf_Load_ODS_SALES_From_OLTP_xx and select View Persistent Values.
a)

The value of $$WORKFLOW_RUNS should be 1.

Figure 127: Value of $$WORKFLOW_RUNS after the first run

4)

Run the workflow six more times to simulate a week's normal runs.

5)

After the last run, how is the Gantt chart different?

Extra Credit:
Modify the Workflow to fail if any of the Sessions in the Workflow fail.
Hint: You will need to use more than one Control task.
Hint: You can force a Session failure by changing to a Relational connection that references a
database schema that does not have the table in it. For example, change the target table to use
Relational connection OLTP.

PowerCenter 8.x Level I Developer

Lab 12-1

127

Answers
7.5. After the last run, how is the Gantt char different?
The second Session task is shown connected to the Decision task, and has a status of Succeeded.

128

Lab 12-1

PowerCenter 8.x Level I Developer

Lab 12-2: Using Events and Timers

Scenario:
The sales summary tables can only be loaded after the ODS_SALES table is loaded.
Management has further determined that if the workflow takes more than 15 minutes to
run, you should fail the workflow as something has probably gone wrong.
You created a Mapping, (m7_Load_SALES_SUMMARIES), that loads the Sales
Summary tables.
You created a Mapping (m6_Load_ODS_Sales) that loads the ODS_Sales table.
Data must be retrieved from and written to the ODS database schema.
Goals:

Create a Workflow that loads the ODS_SALES table, then raises an User-Defined event.
Wait for the User-Defined event, then load the Sales Summaries tables.
Stop the workflow nicely if the Sales Summary tables load properly.
Create a third branch to the workflow that starts a 15-minute timer. If the time limit is
reached, then fail the workflow.
Set the workflow to run at a particular time.
Duration:
35 minutes

Instructions
Step 1.
1)

Make a Session Reusable

In the Workflow Manager application, locate the Workflow in your folder named
wkf_Load_SALES_SUMMARIES_xx and drag it into the Workflow Designer
workspace.
a)

Edit the Session

(i)

In the General tab, check the "Make reusable" checkbox.

(ii) Click Yes to make the Session reusable.

(iii) Click OK.
b)

Step 2.

Save the Workflow.

Create and Configure the Workflow

1)

Create a new Workflow named wkf_Load_ODS_SALES_SUMMARIES

2)

In the Create Workflow dialogue, select the Events tab.

PowerCenter 8.x Level I Developer

Create a new Event named Load_ODS_SALES_IS_DONE

Lab 12-2

129

Figure 128: Create the Workflow Event

3)

Drag the Session s_m8_Load_ODS_Sales into the workflow and link the Start task
to it.

4)

Save your work.

Step 3.

Create the Event Raise Task

1)

Use the Event Raise Task button (

2)

Link the Session task to the Event Raise task.

3)

Double-click the link to edit it.

4)

Add a link condition to ensure that the Event Raise task executes only if the Session
task was successful.
a)

) to add an Event Raise task to the Workflow.

Double-click the pre-defined function "Status" and set the condition so that the
status must be SUCCEEDED. (See figure.)

Figure 129: Link Condition

5)

130

Edit the Event Raise task.

a)

Change its name to er_Load_ODS_SALES_IS_DONE.

b)

In the Properties tab, set the User Defined Event to wait for. (See figure.)

Lab 12-2

PowerCenter 8.x Level I Developer

Figure 130: Event Wait Task Expression

6)

Save your work.

Step 4.

Create the Event Wait Task

1)

Use the Event Wait task button (

2)

Link the Start task to the Event Wait task.

3)

Edit the Event Wait task

a)
b)

) to add an Event Wait task to the Workflow.

Rename it ew_Load_ODS_SALES_IS_DONE.
In the Events tab, set a User-Defined event which the Event Wait task will wait
for before executing. See the figure.

Figure 131: Event Wait Task User-Defined Event

4)

Step 5.

Save your work.

Add a Second Session

1)

Drag the session s_m7_Sales_Summaries_xx into the Workflow.

2)

Link ew_Load_ODS_SALES_IS_DONE to it.

3)

Double-click the link.

PowerCenter 8.x Level I Developer

Lab 12-2

131

a)

Add a link condition that checks whether the Event Wait task has completed
successfully. (See figure.)

Figure 132: Link Condition

Step 6.

Create the first Control Task

1)

Use the control task button (

2)

Link s_m7_Sales_Summaries to the Control task.

3)

Add a Link condition that checks whether s_m7_Sales_Summaries has returned a

status of SUCCEEDED.

4)

Edit the Control task.

a)
b)

) to add a Control task to the workflow.

Rename it ctl_Stop_Workflow_Nicely
In the Properties tab tell the Control task to stop the top-level workflow (see
example below).

Figure 133: Control task Properties

c)

132

Click OK.

Lab 12-2

PowerCenter 8.x Level I Developer

Step 7.

Add a Timer task

1)

Use the Timer task button (

2)

Link the Start task to it.

3)

Edit the Timer task.

a)
b)

) to add a Timer task to the workflow.

Rename it tmr_Wait_15_Minutes
In the Timer tab tell the Timer task to count 15 minutes from the time the
parent workflow started (see example below).

Figure 134: Timer task Timer settings

c)

Step 8.

Click OK

Create the Second Control Task

1)

Use the control task button (

2)

Link tmr_Wait_15_Minutes to the Control task.

3)

Add a Link condition that checks whether tmr_Wait_15_Minutes has returned a

status of SUCCEEDED.

4)

Edit the Control task.

a)
b)

) to add a second Control task to the workflow.

Rename it ctl_Fail_Load_ODS_SALES_SUMMARIES
In the Properties tab tell the Control task to Fail the top-level workflow (see
example below).

PowerCenter 8.x Level I Developer

Lab 12-2

133

Figure 135: Control task Properties

c)

Step 9.

Click OK.

Verify the workflow

1)

Save your work.

2)

Verify the workflow is valid

3)

Right-click in the workspace and select Arrange Horizontal

Figure 136: wkf_Load_ODS_SALES_SUMMARIES completed

Step 10.

134

Use the Scheduler to set the Workflow to Start at a Given time

1)

Edit the workflow

2)

Select the Scheduler tab

3)

Click the Scheduler button (

4)

In the Edit Scheduler windows, select the Schedule tab

5)

Check the Run on Integration Service Initialization check box

6)

Set the workflow to start a few minutes from now. For example, if it is 12:55AM,
set the workflow to start at 1:00AM. (see figure below)

Lab 12-2

PowerCenter 8.x Level I Developer

Figure 137: Edit Scheduler window set to start the workflow a few minutes from now

7)

Click OK

8)

Click OK

9)

Save your work

PowerCenter 8.x Level I Developer

Lab 12-2

135

Step 11.

Start the Workflow and Monitor the Results

1)

The workflow will start, but not execute until the date and time set in the Scheduler.
Wait until it starts.

2)

Review the Workflow results in the Gantt view of the Workflow Monitor. It should
appear similar to the figure below:

Figure 138: Gantt Chart of the Completed Workflow Run

Note that the first Control task stopped the workflow before the second one failed it.
The first Control task is needed so the second one doesn't execute after 15 minutes
every time the workflow is run.

136

Lab 12-2

PowerCenter 8.x Level I Developer

Workshop 1: Mapping Design Workshop

What to Consider
The mapping process requires much more up-front research than it may appear. Before
designing a Mapping, it is important to have a clear picture of the end-to-end processes that
the data will flow through.
Design a high-level view of the mapping and document a picture of the process with the
mapping, using a textual description to explain exactly what the Mapping is supposed to
accomplish and the methods or steps it will follow to accomplish its goal.
Figure 139: Example of a high-level process overview

After the high-level flow has been established, document the details at the field level,
listing each of the Target fields and the Source field(s) used to create each Target field.
o

Document any expression that may be needed to generate the Target field (e.g.: a
sum of a field, a multiplication of two fields, a comparison of two fields, etc.).

Whatever the rules, be sure to document them at this point, and remember to
keep it at a physical level.

The designer may have to do some investigation at this point for some business
rules. For example, the business rules may say "For active customers, calculate a
late fee rate." The designer of the Mapping must determine that, on a physical
level, this translates to "for customers with an ACTIVE_FLAG of 1, multiply the
DAYS_LATE field by the LATE_DAY_RATE field."

Create an inventory of Mappings and reusable objects. This list is a "work in progress"
and will have to be continually updated as the project moves forward.
o

The administrator or lead developer should gather all the potential Sources, Targets, and
reusable objects and place them in a folder accessible to all who may need access to
them.
o

These lists are valuable to everyone, but especially for the lead developer. These
objects can be assigned to individual developers and progress tracked over the
course of the project.

In our labs, this has been SC_DATA_STRUCTURES.

If a shared folder for Sources and Targets is not available, the developer will need to
obtain the Source and Target database schema owners, passwords, and connect strings.

PowerCenter 8.x Level I Developer

Workshop 1

137

With this information, ODBC connections can be created in the Designer tool to
allow access to the Source and Target definitions.

This information will also be needed to create Connections in the Workflow

Manager.

Reusable objects need to be properly documented to make it easier for other developers
to determine whether they can/should use them in their own development.

A developer's specifications for a Mapping should include, at a minimum, the required

Sources and Targets, and any additional information regarding derived ports, and how
the ports relate from the Source to the Target.

The Informatica Velocity methodology provides a matrix that assists in detailing the
relationships between Source fields and Target fields (Mapping Specifications.doc). It
also depicts fields that are derived from values in the Source and eventually linked to
ports in the Target.

Document any other information about the Mapping that is likely to be helpful in
developing it. Helpful information may, for example, include Source and Target
connection information, Lookups (and how to match data in the Lookup table),
potential data issues at a field level, any known issues with particular fields, pre-or postMapping processing requirements, and any information about specific error handling
requirements for the Mapping.

The completed Mapping design should then be reviewed with one or more team
members for completeness and adherence to the business requirements.
o

In addition, the design document should be updated whenever the business rules
change, or if more information is gathered during the build process.

Mapping Specifics
The following tips will make the Mapping development process more efficient. (Not in any
particular order.)
One of the first things to do is to bring all required Source and Target objects into the
Mapping.

138

Only connect fields that are needed or will be used.

o

Note, however, that all ports must be connected from the Source definition to the
Source Qualifier transformation.

Only needed fields should be projected from Source Qualifiers that originate with
Relational tables. The SQL that PowerCenter generates will include only the
needed fields, reducing computing resource requirements. In this case, only
connect from the Source Qualifier those fields that will be used subsequently.

Filter rows early and often. Only manipulate data that needs to be moved and
transformed. Reduce the number of non-essential records passed through the Mapping.

Decide if a Source Qualifier join will net the result needed, versus creating a Lookup to
retrieve desired results.

Reduce the total number of transformations. Excessive number of transformations will

increase overhead.

Workshop 1

PowerCenter 8.x Level I Developer

Make use of variables (local or global) to reduce the number of times functions will have
to be used.

Watch your datatypes. The Informatica engine converts compatible datatypes

automatically, but unnecessary conversion is inefficient.

Make use of variables, reusable transformations, and Mapplets as "reusable code." These
will leverage the work being done by others, promote standardization, and ease
maintenance tasks.

Use active transformations early in the process to reduce the number of records as early
in the Mapping as possible.

When joining Sources, select the appropriate master (driving) table.

Utilize single-pass reads. Design Mappings to utilize one Source Qualifier to populate
multiple Targets.

Remove or reduce field-level stored procedures.

o

Even though PowerCenter does a lot to increase efficiency, Stored Procedure

objects will be executed for each record, and slow performance.

Lookup transformation tips:

o

When the Source is large, cache lookup tables columns for lookup tables with
500,000 rows or less on 32-bit platforms with limited RAM memory.

Standard rule of thumb is not to cache tables over 500,000 rows on 32-bit
platforms with limited RAM.

Use equality (=) conditions if possible in the Condition tab.

Use IIF or DECODE functions when lookup returns small rowsets.

Avoid data comparisons in lookup; convert to string.

Operations and Expression transformation tips

o

Numeric operations are faster than string.

Trim Char and Varchar fields before performing comparisons.

Operators are faster than functions (i.e., || is faster than the CONCAT function).

Use flat files. File read/writes are faster than database read/writes on the same
server. Fixed-width files are faster than delimited file processing.

PowerCenter 8.x Level I Developer

Workshop 1

139

Workshop
Scenario:

Management wants the ability to analyze how certain promotions are performing. They
want to be able to gather the promotions by day for each dealership, for each product
sold.

Goals:

Design and create a mapping to load the aggregate table.

Duration:
120 minutes

Instructions
Sources and Targets
Sources: TRANSACTIONS and PRODUCT_COST
These relational tables contain sales transactions and Product cost data for seven days. They
are located in the SDBU schema. For the purpose of this mapping, we will read all the data
in these tables.
These tables can be joined on PRODUCT_ID and PRODUCT_CODE
Figure 140: TRANSACTION table definition

Figure 141: PRODUCT_COST table definition

Target: ODS_PROMOTIONS_DAILY
This is a relational table located in the ODSxx schema. After running the Mapping, it should
contain 1283 rows.

140

Workshop 1

PowerCenter 8.x Level I Developer

Figure 142: ODS_PROMOTIONS_DAILY table definition

Mapping Details
In order to successfully create the mapping, you will need to know some additional details.
Management has decided that they don't need to keep track of the Manager Discount
and the Employee Discount (PROMO_ID 105 and 200), so these will need to be
excluded from the load.

The PRODUCT_DESC can be obtained from the PRODUCT table by matching on

the PRODUCT_ID.

The DATE_DESC can be obtained from the STG_DATES table by matching the
TRANSACTION table DATE_ID to the DATE_ID in STG_DATES.

UNITS_SOLD is derived by summing QUANTITY.

REVENUE is derived by taking the value in the QUANTITY port times the
SELLING_PRICE and then subtracting the DISCOUNT, HOLDBACK and
REBATE.

Most of the discounts are valid but occasionally they may be higher than the
acceptable value of 17.25%. When this occurs you will need to obtain an acceptable
value based on the PROMO_ID. The acceptable value can be obtained from the
PROMOTIONS table by matching the PROMO_ID.

The DISCOUNT is a percentage stored as a number. To calculate the actual

discount in dollars, divide the DISCOUNT by 100 and multiply it by the
SELLING_PRICE.

Revenue per unit is REVENUE divided by QUANTITY.

COST is derived by summing UNIT_COST.

COST_PER_UNIT is derived by summing the UNIT_COST and dividing it by the

sum of QUANTITY.

Data will be aggregated by PROMO_ID, DEALERSHIP_ID, PRODUCT _DESC,

and DATE_DESC.

Save your work often!!!

PowerCenter 8.x Level I Developer

Workshop 1

141

SOURCE TO TARGET FIELD MATRIX

Target
Table

142

Target
Column

Source

File

Source
Column

Workshop 1

Expression

Default Value for

Nulls

PowerCenter 8.x Level I Developer

Run Details
Your Task Details, Source/Target Statistics, and preview of the Target data should be similar
to the figures below.
Figure 143: Task Details of the Completed Run

Figure 144: Source/Target Statistics for the completed run

Figure 145: Data Preview of ODS_PROMOTIONS_DAILY

PowerCenter 8.x Level I Developer

Workshop 1

143

144

Workshop 1

PowerCenter 8.x Level I Developer

Workshop 2: Workflow Design Workshop

What to Consider
The Workflow process requires some up-front research. Before designing a Workflow, it is
important to have a clear picture of the task-to-task processes.
Design a high-level view of the Workflow and document the process within it, using a
textual description to explain exactly what it is supposed to accomplish and the steps or
methods it will use to accomplish its goal.
Figure 146: Example of High-Level Workflow Overview

The load development process involves the following steps:

Clearly define and document all dependencies.

Analyze the processing resources available.

Develop operational requirements.

Develop tasks, Worklets, and Workflows based on the results.

Create an inventory of Worklets and reusable tasks. This list is a "work in progress" list
and will have to be continually updated as the project moves forward.

The lists are valuable to everyone, but particularly for the lead developer.

Making an up-front decision to make all Session, Email and Command tasks
reusable will make this easier.

The administrator or lead developer should put together a list of database connections to
be used for Source and Target connection values.

Reusable tasks must be properly documented to make it easier for other developers to
determine whether they can or should use them in their own development.

PowerCenter 8.x Level I Developer

Workshop 2

145

If the volume of data is sufficiently low for the available hardware to handle, you may
consider volume analysis optional, developing the load process solely on the dependency
analysis.

Another possible component to add into the load process is sending email. Three email
options are available for notification during the load process:

Post-session emails can be sent after a Session completes successfully or when if fails.

Email tasks can be placed in Workflows before or after an event or series of events.

Emails can be sent when Workflows are suspended.

Document any other information about the Workflow that is likely to be helpful in
developing. Helpful information may, for example, include Source and Target database
connection information; pre- or post-Workflow processing requirements; and any
information about specific error handling for the Workflow.

Create a Load Dependency analysis. This should list all Sessions by dependency, along
with all other events (Informatica or other) they depend on.

Also, be sure to specify the dependency relationship between each Session or event,
the algorithm or logic needed to test the dependency during execution, and the
impact of any possible dependency test results (e.g., don't run a Session, fail a
Session, fail a parent or Worklet, etc.)

Create a Load Volume analysis. This should list all the Sources and row counts and row
widths expected for each Session.

146

If the hardware is not adequate to run the Sessions concurrently, you will need to
prioritize them. The highest priority within a group is usually assigned to Sessions
with the most child dependencies.

This should include all Lookup transformations in addition to the extract Sources.
The amount of data that is read to initialize a Lookup cache can materially affect the
initialization and execution time of a Session.

The completed Workflow design should then be reviewed with one or more team
members for completeness and adherence to the business requirements.

The design document should be updated whenever the business rules change, or if more
information is gathered during the build process.

Workshop 2

PowerCenter 8.x Level I Developer

Workflow Specifics
The following tips will make the Workflow development process more efficient (not in any
particular order).
When developing a sequential Workflow, use the Workflow Wizard to create Sessions in
sequence. You also have the option to create dependencies between Sessions.

Use a parameter file to define the values for parameters and variables used in a
Workflow, Worklet, Mapping, or Session. A parameter file can be created using a text
editor such as WordPad or Notepad. List the parameters or variables and their values in
the parameter file. The use of Parameter files is covered in the Level 2 Developer course.
Parameter files can contain the following types of parameters and variables:

Workflow variables

Worklet variables

Session parameters

Mapping parameters and variables

When using parameters or variables in a Workflow, Worklet, Mapping or Session, the

Integration Service checks the parameter file to determine the start value of the
parameter or variable. Use a parameter file to initialize Workflow variables, Worklet
variables, Mapping parameters and Mapping variables. If not defining start values for
these parameters and variables, the Integration Service checks for the start value of the
parameter or variable in other places.

Session parameters must be defined in a parameter file. Since Session parameters do not
have default values, when the Integration Service cannot locate the value of a Session
parameter in the parameter file, it fails to initialize the Session.

To include parameter or variable information for more than one Workflow, Worklet, or
Session in a single parameter file, create separate sections for each object within the
parameter file.

You can create multiple parameter fields for a single Workflow, Worklet, or Session and
change the file these tasks use as necessary. To specify the parameter file the Integration
Service uses with a Workflow, Worklet, or Session, do either of the following:

Enter the parameter file name and directory in the Workflow, Worklet, or Session
properties.

Start the Workflow, Worklet or Session using pmcmd and enter the parameter
filename and directory on the command line.

On hardware systems that are underutilized, you may be able to improve performance by
processing partitioned datasets in parallel in multiple threads of the same Session
instance running on the Integration Service node.

However, parallel execution may impair performance on over-utilized systems or

systems with smaller I/O capacity.

PowerCenter 8.x Level I Developer

Workshop 2

147

148

Incremental aggregation is useful for applying captured changes in the Source to

aggregate calculations in a Session. If the Source changes only incrementally, and you
can capture changes, you can configure the Session to process only those changes.

This allows the Integration Service to update your Target incrementally, rather than
forcing it to process the entire Source and recalculate the same calculations each
time you run the Session.

This is particularly useful for "slowly-changing dimension" tables. Slowly changing

dimensions are covered in the Level 2 Developer class.

Target Load Based Strategies:

Loading directly into the Target is possible when the data is going to be bulk loaded.

Load into flat files and bulk load using an external loader.

Load into a mirror database.

From the Workflow Manager Tools menu, select Options and deselect the option to
"Show full names of task." This will show the entire name of all tasks in the Workflow.

Workshop 2

PowerCenter 8.x Level I Developer

Workshop
Scenario:

All the staging tables need to be loaded in a single Workflow.

Goals:

Design and create a Workflow to load all of the staging tables.

Duration:
120 minutes

Instructions
Mappings Required
This section contains a listing of the Mappings that will be used in the workflow:
m_Load_STG_PAYMENT_TYPE

m_Load_STG_Product

m_Load_STG_Dealership

m_Load_STG_PROMOTIONS

m_Load_STG_CUSTOMERS

m_Load_STG_TRANSACTIONS

m_Load_STG_EMPLOYEES

For your convenience, reusable Sessions have been created for these mappings. You can
COPY them from the SC_DATA_STRUCTURES folder to your folder. (One or more of
these Sessions may already be in your Sessions subfolder.) Remember to use the Repository
Manager to copy the sessions. If the copy wizard asks to resolve any conflicts, tell it to replace
old definitions with new ones.
The names of the sessions are:

s_m_Load_STG_PAYMENT_TYPE

s_m_Load_STG_PRODUCT

s_m_Load_STG_DEALERSHIP

s_m_Load_STG_PROMOTIONS

s_m_Load_STG_CUSTOMERS

s_m_Load_STG_TRANSACTIONS

s_m_Load_STG_EMPLOYEES

PowerCenter 8.x Level I Developer

Workshop 2

149

Workflow Details
1.

Name the workflow wkf_LOAD_ALL_STAGING_TABLES_xx.

The Workflow needs to start at a certain time each day. For this workshop, set the
start time to a couple of minutes after you complete the Workflow. Remember that
the start time is relative to the time on the Integration Service process machine, not
a developer's local machine.

2.

No Session can begin until an indicator file shows up. The indicator file will be
named fileindxx.txt, and will be created by you using a text editor. You will need to
place this file in the directory indicated by the Instructor after you start the
Workflow. (If you are in a UNIX environment, you can skip this requirement.)

3.

In order to utilize the CPU in a more efficient manner, you will want to run some of
the Sessions sequentially and some of them concurrently.
a.

The Sessions s_m_Load_STG_PAYMENT_TYPE,

s_m_Load_STG_PRODUCT, s_m_STG_DEALERSHIP and
s_m_Load_STG_PROMOTIONS can all be run sequentially.

b.

The Session s_m_Load_STG_CUSTOMERS can be run concurrently with

the Sessions in point a.

c.

If any of these sessions fail, an email should be sent to the administrator,

and the Workflow aborted. Use [email protected] as the email user
name.

d.

The session containing Mapping m_STG_EMPLOYEES should be run

only after the five previously-mentioned Sessions complete successfully.

e.

The Session s_m_Load_STG_TRANSACTIONS can be run concurrently

with the session s_m_STG_EMPLOYEES.

f.

If either of these Sessions fails, an email should be sent to the

Administrator.

4.

All Sessions truncate the Target tables and should be pointed to the correct
relational database connections.

5.

Management wants the Workflow to run for a maximum of 50 minutes. If the

Workflow takes longer than 50 minutes, an email must be sent to the administrator.
If the Workflow completes on time, the email should not be sent.
This is a subtle point. It means you must stop the workflow nicely before the timer
task has a chance to finish running and the "long running workflow" email gets sent.
Consider which task to use and where to place it in the workflow so that when all
other tasks run to completion properly, the "long running workflow" email is not
sent.

Final Point
More than one solution is possible. You will know that your solution has worked when all
the Sessions complete successfully.

150

Workshop 2

PowerCenter 8.x Level I Developer