Discussion Forum

Back to topics

Handling Inventory Management Scenario using Snapshot method

Subscribe to topic

28-May-10 08:52

Message # 350671

Sai Ram

Hi Experts,

I have worked on various SAP BW projects but never got chance to work on Inventory management. Now for the first time I am responsible for implementing the Inventory management scenario to our client and need your help.

I am quite confused of these 2 scenarios(non cumulative and snapshot). We are asked to implement both the scenarios is parallel. i.e. Daily reports need to be done from 0IC_C03 and Monthly reports from snapshot.
We are using daily cube IC_C03 for detialed report and Snapshot for high level....

The following are the new infoobjects we need to add them in detailed cube(IC_C03) for more drill down...

0WHSE_NUMWarehouse Number/Warehouse Complex
0COMP_CODECompany code
0CO_AREAControlling Area
0COSTCENTERCost Center
0COSTOBJCost Object
0COORDEROrder Number
0PROFIT_CTRProfit Center
0ACCOUNTG/L Account Number
0OI_EBELNDocument Number of Purchasing Document
0DOC_NUMMaterial Document Number
0OI_EBELPItem Number of Purchasing Document
0DOC_ITEMItem in Material Document
0DOC_DATEDocument Date in Document
0PSTNG_DATEPosting Date in the Document
0DOC_NUMBERSales Order Number
0SCHED_LINESchedule Line Sales Order
0S_ORD_ITEMItem Number in Sales Order
0SOLD_TOCustomer’s Account Number
0MAT_PLANTMaterial Plant
0MAT_SALESMaterial Sales
ZGRUNDReason for goods movment
ZKZBEWMovement Indicator

Just wanted to know/Clarification on the following :

1. Do we need all these newly added object at snapshot cube or DSO as these are not in my high level report requirements.
2. What is the purpose of ZST_movements explained in one blog(i.e.Implementing an Optimal Stock Reporting Model) of this portal, how can that be useful to me.
3.What is the loading procedure that need to be followed for both scenarios, it was explained in blog with screen design but I was quite confused with it.
4. what would be the key fields of DSO.

5. Why are the characteristics of Snapshot cube in (SAP pdf "How to handle inventory scenarios") are all Line item dimensions.

Please clarify the above requirements. Thanks in advance....
Let me know for any more information.

thanks

SaiRam

31-May-10 14:30

Reply # 351775 on 350671

Sergei Peleshuk (administrator)

hello Sai,

Let me answer your questions. I presume there may be more than one opinion on this so other inputs are welcome as well.

1. Snapshot view objects are usually used for monthly or aggregated reporting. I suggest to do a thorough analysis on what kind of reports you will have to do at this level and include only those objects in the Snapshot DSO & Cube that are need for this kind of reporting.

2. ZST_MOVEMENTS is an example of a generic extractor you can build if your source is SAP Retail. This extractor would capture a higher level of detail than the standard ones. However, if you are ok with the level of detail coming from the standard stock extractors I would suggest using them instead.

3. The sequence of loads is a bit tricky as it depends on your reporting requirements. The non-cumulative daily cube is usually loaded directly from PSA for stock movements. In this scenario daily stock balances are not stored in the cube but recalculated at query run time.

In the snapshot view we do aggregations in the DSO based on key fields and then move summarized data to the monthly cube. Also, once a month we have to calculate a stock balance for the month and copy it as an opening balance for the new month. In this scenario monthly stock balances are stored in the monthly cube.

4. Snapshot DSO key fields should have all fields that have to be reported on in the cube. Usually they include: Material, Plant, Movement Type, Reason code, Period. If you plan to do financial reporting from the same cube you want to add GL Account, Cost Center, Profit Center, Company Code.

5. I do not recommend setting any dimension as a line item dimension in the stock cubes unless you plan to include document numbers there. The number of stock transactions stored in the cube usually is far superior to the size of any dimension based on the infoobjects mentioned in the answer 4. And for the performance reasons we have to keep our dimension tables within 20% of the fact table.