Tuesday, December 25, 2012

Transport Process Chains to Test System




Q.What is the best way to transport process chains to test system?
I got many other additional and unwanted objects got collected when I tried for collection of process chains from transport connection.

To transport a process chain the best is to transport only objects created for the process chain. On my system I created specific obejcts for PC : Infopackages, jobs, variant. those objects are only use for PC. By this way I avoid errors when users restart load or job manually. 
So when I want to transport a process chain I go in the monitor and select the PC make a grouping on only necessary objects, and I go through the tree generated to select only what I need. Then I go in SE10 to check if the transport contains not other objects which can impact my target system. 
You can avoid some uncessary objects by clicking in Grouping > Data flow before & Data Flow After . For example you already have infopackages in your target system but not process chains & you only want to transport only process chain without any other objects like transfer structure or infopackages . You can choose before or after option .
 
You can also choose hierachries or display option from the Display tab too if you have objects in bulk but make sure all object are selected ( in case when different-2 process chain having different kind of object then better use Hierarchy, not list ) 
While Creating these TR some objects may be in use or locked in other TR so first release them by Tcode Se03 ,using unclock object ( Expert Tool ). 
These options can reduce your effort while collecting your objects , even after so much effort you get some warning or Error like :- objects are already in system then ask basis to use overwrite mode.

Transport a specific infoobject........

Q.How to transport a specific info object? I tried to change it and then save but the transport request won't appear. How to manually transport that object?

1. Administrator Workbench (RSA1), then Transport Connection
2. Object Types, then from the list of objects put the requested one on the right side of the screen (drag & drop)
3. Click "Transport Objects", put the development class name and specify the transport (or create the new one)
4. Transaction SE01, check transport and release it
5. Move the transport up to the another system.
 
If you change and reactivate the infoobject, but get no transport request, this means that your infoobject is still in $tmp class.
go in to the maintenance of the infoobject, menu extras, object directory entry and change the development class. at this point you should have a pop-up requesting a transport request
 
If you're not getting a transport request when you change and activate, it could also be that the InfoObject is already on an open transport. 
When you collect the object in the transport connection as described above, you will see in the right hand pane an entry called Transport Request. If there is an entry here, the object is already on a transport and this gives you the transport number. 
You can then use SE01 or SE10 to delete the object from the existing transport if that is what you want to do then, when you change and activate the object again, you should be prompted for a transport request. Alternatively, you can use the existing transport depending on what else is on it.

Q.How To Do Transports in BW?

Step by step procedure for transporting in BW:
1. In RSA1 go to Transport Connection
2. Select Object Types Your Object that you want to transfer.
3. Choose grouping method (in data flow before and after)
4. Drag and drop your object.
5. Click the Truck button to transfer
6. In Source System (e.g Dev SE09).
    a. Expand to choose your child request
    b. Click on the release button (truck)
    c. Choose the parent request and click the Truck button release.
7. In Target System (e.g QA) go to STMS
    a. Click on Truck button (Import Overview)
    b. Dbl click on your QA system queue
    c. Clck on Refresh
    d. Clk on adjust import queue
    e. Select ur request and click on F11.
                            
Q.Is it possible to copy a process chain in BW 3.1? If so, how?

In RSPC, double click the process chain so that you can see it in the left hand pane. In the box where you type in the transaction code, type COPY and hit Enter.

Q.Infocube Compression.

I was dealing with the tab "compression" while managing the infocube, was able to compress the infocube and send in the E- table but was unable to find the concrete answer on the following isssues:

1. What is the exact scenario when we use compression?

Compression creates a new cube that has consolidated and summed duplicate information.



2. What actually happens in the practical scenario when we do compression?

When you compress, BW does a group by on dimensions and a sum on measures... this eliminates redundent
information.

3. What are the advantages of compressing a infocube?

Compressed infocubes require less storage space and are faster for retrieval of information.

4. What are the disadvantages of compressing a infocube?

Once a cube is compressed, you cannot alter the information in it. This can be a big problem if there
is an error in some of the data that has been compressed.

Q.I understand the advantage to compressed the infocube is the performance.  But I have a doubt.  If I compressed one or more request ID of my infocube the data it will continue to appear in my reports (Analyzer)?

The data will always be there in the Infocube. The only thing that would be missing is the request id's.. you can take a look in to your packet dimension and see that it would be empty after you compress.
Compression yeap its for performance. But before doing this compression you should keep in mind one thing very carefully.
1) If your cube is loading data with custom defined deltas you should check whether delta is happening properly or not, procedure is compress some req and schedule the delta.
2) If your system having outbounds from cube and this happening with request ids then you need to follow some other procedure because request ids wont be available after compression.
These two things are very important when you go for compression.