Loading data from Cube to Planning area

[Former Member]

Hi,

If I am loading data from a cube to a planning area using transaction TSCUBE,

does the system load data into planning area for the combinations that exist in the cube or does it load for all CVCs?

For example,

I have my CVC as Plant, Material, Customer

If there are 4 CVCs in the POS that were previously generated as

Plant--Material--Customer

01--

---

M1--

---

C1

01--

---

M2--

---

C3

01--

---

M2--

---

C2

01--

---

M4--

---

C5

If the cube has data like this:

Plant--MaterialCustomer----Qty.

01--

---

M1

---

C1--

---

10

01--

---

M2

---

C3--

---

20

01--

---

M2

---

C2--

---

5

(doesnot have the last combination), then if I use TSCUBE transaction to load data to Planning area from this cube,

is the data loaded as

Plant--MaterialCustomer----Qty.

01--

---

M1

---

C1--

---

10

01--

---

M2

---

C3--

---

20

01--

---

M2

---

C2--

---

5

Only for the 3 combinations that exist in the cube and not load anything for the last one

OR

is the data loaded as

Plant--MaterialCustomer----Qty.

01--

---

M1

---

C1--

---

10

01--

---

M2

---

C3--

---

20

01--

---

M2

---

C2--

---

5

01--

---

M4

---

C5--

---

0

Load all 4 combinations and send 0 as the cube doesnot have this combination?

Hope I am clear on this question.

Thanks.

[Former Member · Accepted solution]

Hi,

The data will be loaded only for the CVC's existinig in the POS and it will return a warning for the combination non-existent in the POS.

Regards

Vinod

[Former Member · Accepted solution]

Thanks for the reply. I have always seen messages the other way. If there is a combination in the cube that is not in the planning object structure, then it gives you a message but has never seen messages the way you said....may be I have to test the scenario.

Thanks any way for your time.

[srinivas_krishnamoorthy · Accepted solution]

I would tend to go with the former. Zero records are NOT passed if the CVC does not exist in Infocube. In many designs I have seen Infocube to be pass through entities passing sections of data onto POS. Any data update to POS should not wipe out existing data in POS for CVCs not existing in Infocube but existing in POS. Hence only CVCs present in Infocube will be passed onto POS.

This feature has to be present and is present by design in SAP otherwise imagine loading the complete list of CVCs everytime leading to a grossly inefficient process.

[Former Member · Accepted solution]

just to extend the discussion

this cube to PA copy is done by breaking down the time buckets

so if even one combination of 01--

---

M4--

---

C5 for any time bucket in the cube has a entry (blank or zero included), all the other values in the PA for the combination will be overwritten with a zero

unless, you are using TSCUBE in SCM 5.0 (and probably 4.x) you have an option of additional settings through which you will not be able to replace non-zero values with zero values from the cube

also if you want to change this behavior, there is a badi called /SAPAPO/SDP_IC_LOAD

it has methods CHANGE_CUBE_DATA and CHANGE_DATA_BEFORE_SAVE that can be used to manipulate the cube data

[Former Member · Accepted solution]

Thanks a lot Vinod, Srinivas and Harish. The reason why I am asking you is that we have a scenario where we get this situation.

We initially get data from R/3 to BW to APO like the below:

Plant--MaterialCustomer----Qty.

01--

---

M1

---

C1--

---

10

Later when the customer is changed or bought out by somebody C1 is changed to C2. Some times when the business doesnot know who the customer is initially they just put C1 as dummy and then after sometime replace it by C2. Then the new record coming in is as follows:

Plant--MaterialCustomer----Qty.

01--

---

M1

---

C2--

---

10

BW can identify changes in transaction data but not in Master data. What I mean by this is when Qty. 10 changes from 10 to 20, the system can identify it in deltas.

If the customer (master data) changes to C2 from C1, the system thinks it's a new record all together then if I use delta loads, it gets me the following:

Plant--MaterialCustomer----Qty.

01--

---

M1

---

C1--

---

10

01--

---

M1

---

C2--

---

10

If I am looking at Plant and Material Level, my data is doubled.

So we are planning to do a full load that works like this:

1. Initial data like the below:

Plant--MaterialCustomer----Qty.

01--

---

M1

---

C1--

---

10

The CVC is created and the planning area has Qty.10

Then we delete the contents of cube and do a full load into the cube with changed customer

Plant--MaterialCustomer----Qty.

01--

---

M1

---

C2--

---

10

This time a new CVC is created. Then we have another 10 loaded into Planning area.

If the system loads all CVCs, then the it would send

Plant--MaterialCustomer----Qty.

01--

---

M1

---

C1--

---

0

01--

---

M1

---

C1--

---

10

If the system loads only combinations in cube,

then it loads

Plant--MaterialCustomer----Qty.

01--

---

M1

---

C2--

---

10

But the system already has another 10 for Customer C1 duplicating the values.

We are trouble in the second case.

We had to go fr this solution instead of realignment as our business has no way pf knowing that C1 was replaced by C2.

Hope I am clear.

[Former Member · Accepted solution]

Hi,

As mentioned by others previously if the CC is not in your cube, it will not modify the data in your planning area.

If I understand your issue correctly, you should "clean" your planning area before reloading the data from the cube: for example run a macro that set all the records to 0. Then reload your data from the cube.

Please note that this will not delete the CC from your planning area. So you might end up with a lot a CC that are not used anymore...

Good luck

Julien