Data Validation after Upload

This section will outline the procedures for doing a “spot-check” of the data integrity for each subject area.  This is high level testing that should be performed on a regular basis after transaction data loads.

For Data validation please refer to Integration test scripts documents in

Y:\BW\Dw - Business Information Warehouse\PSI - Phase 1\3 - BW Realization\Integration Test Scripts

Problem Resolution

This section will define some of the more common problems you may encounter in supporting BW, as well as their solutions.

Data Load Failure due to Invalid Characters

Sony Latin America has characters within the master data that are outside of those allowed within the BW System by default.  This is largely due to several languages being used in the SAP R/3 Source System.  As long as new master data is being created in R/3, there will always be the possibility of a job erring out due to these “invalid” characters.  A typical error message in the Status and Details tabs of the monitor will say,

“Data record 41: Invalid character(s) in 'Ñ ' (InfoObject 0PYMT_METH)”.

To include invalid characters in the list of permitted characters, go to BW Transaction Code RSKC:

Next, copy the entire error message into the Permitted Extra Characters field and then delete it, except for the invalid character itself (this works well if your keyboard cannot type the invalid character):

Click the “Execute” icon to save the Permitted Character.

Now you are ready to try running your job again.

Note:   Do not delete the other Permitted Characters in this field, just continue to add them to the list as errors occur.

Data Load Failure due to Locked InfoObject

You can inadvertently lock an InfoObject from being loaded, if you try to re-load it again after a previous failed job.  In this case, you will see the message in the Details tab of the monitor,

“No access permission as table is already locked”

To unlock the table, click the “Lock Overview” icon:

The locked InfoObject would be displayed here.  To delete the lock, simple right-click on the InfoObject and select “Delete Lock.”  Keep hitting the “Refresh” button until the screen is displayed clear as above.

Refresh

 

Unable to see Data when Executing a Query

An end-user may report that he/she cannot see any data whenever they try to execute a query.  This is more than likely caused by a failed transaction data load that has not been deleted from the InfoCube. 

In order to view any data for an InfoCube, all of the requests must have OK status indicated by a Green Light:

If there is a failed request in the InfoCube, it must be deleted.  In order to check whether a failed request is in an InfoCube, go to the “Manage InfoCubes” transaction in the Admin Workbench:

Right-click on the InfoCube and select “Manage”

OR click the “Active” icon to the right of the InfoCube:

The request with the Red Status indicates a failed load.  The data for this InfoCube is currently not reportable because of this “bad” request.  It needs to be deleted.

To delete the failed request, select the row and then press the “Delete” button.

The trashcan icon will denote that the job is currently running to delete the data.  Keep hitting the “Refresh” icon until it disappears.

Refresh

 

Once the job has completed, data within the InfoCube should be reportable again.

Transporting Query Revisions

While the majority of BW Objects can be transported using the Transport Connection Functionality, the process is slightly different for Queries in that you first have to create the Transport Request, make your changes to the query, and THEN use the Transport Connection functionality.  This process is outlined below:

1)      ALWAYS make the query changes in the SW1, BW Sandbox environment first.

2)      In Development (DW1), create the Transport Request for the Query you are modifying.

Go to the Transport Connection tab in the Administrator Workbench and select the “Request BEx” icon:

Select the “Create” icon to create a new request:

Enter a description for the Transport and then hit the “SAVE” icon. 

Note:     The naming convention for all BW Transports begins with “BW_1_PSI:”

SAVE

 

Note the Transport Number.

3)      Now you can log into the BEx Analyzer, (DW1 environment), and make revisions to the query.  Save your revisions.

4)      You are now ready to use the Transport Connection Functionality to move the revised query through the landscape to Production.

In the Transport Connection tab, make sure that the Grouping is set to “Only Necessary Objects”

Also make sure that the Collection Mode is set to “Collect Automatically”

5)      Select Query Elements and select the Query you want to transport by double-clicking the “Select Objects” icon.

6)      Select the Query and then click the “Copy” icon.

7)      The Query is now in the Installation Window.  Click the “Select All” icon to include all of the Objects in the transport.

8)      Now select the “Transport Objects” icon to generate the Transport Request.

9)      When the selection box pops up, select “Own Requests”

10)  Highlight the Transport Request you created earlier and then click the “Choose” icon.

11)  Confirm that this is the correct transport and then select the “Continue/Enter” icon.

12)  The Transport Request is now generated and a status message will appear at the bottom of the screen indicating that the Transport generated successfully.

Releasing Transports

Now you must release the transport.  Currently, released transports are moved automatically to QW1 every 30 minutes at the top of the hour and half past the hour.  Once your transport is in QW1, ensure that the changes got there OK.  To move the Transport Request to Production, you must notify the Basis Team.

1)      To release a transport, go to BW transaction code SE10.

2)      Your User ID will default in the User field.  Select all of the Request Types and Request Status’ and click the “Display” icon.

3)      The Transport Request status should be “Modifiable”.  You must select the Transport Request Number and then hit the “Release Directly” icon.

Refresh

 

The next screen is for notes and/or documentation.  Just press the “SAVE” icon, which will release the Transport Request.  You can keep hitting the “Refresh” icon until the status is changed to “Released”. 

Now your Transport will automatically move to QW1.

Note:   Lower-Level Tasks of a Transport Request must be released first, before the ‘Main’ Transport can be released.

InfoPackage Groups

An InfoPackage group is a collection of InfoPackages. With the data request, the Scheduler can access InfoPackage groups directly and thereby request more than one InfoPackage at a time. You can edit InfoPackage groups in the InfoSource tree of the Administrator Workbench modeling.  Below is the process for creating InfoPackage Groups.

Creating InfoPackage Groups

1)      In the InfoSources tab of the Admin Workbench, select the “Show InfoPackage Groups” icon.

2)      Now all of the InfoPackage Groups are displayed.  All InfoPackages must be placed within a Group.  To create a new InfoPackage Group, simply right-click on the top-most node and select “Create”.

3)      Enter a description for your InfoPackage Group and select “Enter”.

4)      The new InfoPackage Group is now displayed in the tree.

Note:   You can also create sub-nodes for your InfoPackage Groups by right-clicking on the node where you want it placed and selecting “Create”.

5)     
To add InfoPackages to the InfoPackage Group, simply drag-and-drop the InfoPackage from the InfoSources Tree on the left to the InfoPackage Group on the right.

Note:   Drag-and-Drop the InfoPackage directly on TOP of the InfoPackage Group where you want it to reside.

6)      When you are finished with building your InfoPackage Group, you are ready to schedule it.

Scheduling InfoPackage Groups

The convenience of an InfoPackage Group is that by scheduling one job, you can automatically execute all of the InfoPackages within that InfoPackage Group. You can also monitor the status of the entire InfoPackage Group instead of each individual InfoPackage.  Below is the process for setting up and executing an InfoPackage Group.

You can schedule or view InfoPackage Groups in (2) places:

·         Through the InfoSources tree in the Administrator Workbench in the “Modeling” tab:

·         Through the InfoPackage Group window in the “Monitoring” tab of the Admin Workbench:

1)      Right-Click on the InfoPackage Group and select “Schedule”

2)      You are now in the Scheduler of the InfoPackage Group.  First, you want to check that the left-most column is checked.  This is the flag that tells the system to automatically execute each InfoPackage in the InfoPackage Group.

3)      Next, you have to enter the sequence in which you want the InfoPackages to be executed.  Simply enter the number of the sequence in the “Start Sequence” column.

Note:   Depending on how you have your InfoPackage Groups set up, the sequence of loading the data is important.

1.      Master Data / Attributes

2.      Text

3.      Hierarchies

4.      Transaction Data

4)      Next, you want to flag the “Check” column.  This flag indicates that the system will check whether the request was updated without errors before beginning the next request.

5)      Now you are ready to execute your InfoPackage Group.  Go to the Schedule tab of the InfoPackage Group.  This screen is very similar to the InfoPackage scheduler tab.

6)      You can go to “Selection Options” button and either schedule the job, or start it immediately

7)      Save your selections and then “Execute” the job.

8)      You can go directly into the monitor from this screen, or you can highlight the InfoPackage Group and then select the monitor icon.  The status of each InfoPackage in the InfoPackage Group is displayed.

Scheduling Periodic Jobs in InfoPackages

Scheduling periodic jobs in either an individual InfoPackage or an InfoPackage Group follow the same procedure.  This process is outlined below:

1)      Enter the InfoPackage or InfoPackage Group Scheduler.

2)      In the “Schedule” tab make sure ‘Start Later in Background Processing’ is selected and then click the “Selection Options” button.

3)      Select the “Periodic Job” flag and then select the “Periodic Values” button.

4)     
Select the Period Value for your job by clicking on it (indicated by the brackets) and then selecting “SAVE”.

5)     
Now schedule the Date/Time for your job to begin by pressing the “Date/Time” icon.

Note:      Time entered is in military time.

6)      Now save you entries by pressing the “SAVE” icon.

7)      Make sure the Periodic Processing is set to “Do not Cancel Job”

8)     
Execute the job by selecting the “Start” button.  A status message will appear indicating the job is scheduled.

9)      The clock icon in the InfoSources Tree denotes a job has been scheduled for that InfoPackage / InfoPackage Group.

fangzai33 wrote:

Thanks