Skip to Content
avatar image
Former Member

Modeling Scenario: Inventory management

Good day,

I have been working to implement TD shipments- ( pipe) reporting into HANA.  It requires stock movements; td shipments and their delta; Also caters for different types of units in the system. Previously, I had been able to model the scenario in bi7, now having to do it in hana.

The idea is to pick Oigs/ oigsh/ oigsmvq/ mkpf/mseg/MSEGO1 for the relevant information. I created attribute views for the master data; ( material, plant, storage locations, region) . With the following structure:

Attributes:

  • Plant
  • Material
  • Sender Storage location
  • Receiving Storage location
  • Region
  • Movement type
  • Debit/Credit indicator
  • Posting date
  • Fiscal year
  • Fiscal period
  • stock type

Measures:

  • Quantity loaded
  • Unit of measure
  • Quantity offloaded
  • Unit of measure
  • Quantity in document
  • Unit of measure

Now I have to replicate the delta movements; since there is no 0recordmode in my structure and no facility to utilize delta items from Hana just yet; i am using the debit credit indicator to check material movements; similar to what you see in MB51; cancellations are also treated as negative deductions. Currently i use the universes in the object explorer along with some interval variables to handle certain areas.

Is the right approach? comments please - suggestions are appreciated

Add comment
10|10000 characters needed characters exceeded

  • Get RSS Feed

4 Answers

  • Best Answer
    Mar 19, 2012 at 01:07 PM

    Hi,

    Couple of comments from my side.

    1.You said that you have created Attribute Views.This holds true for master data.As you are combining also the measures,go for Analytic Views(as Neha mentioned).

    2.You can also create a variable in your Analytic View itself,but currently it cant be consumed(say in Universe).Though creating variables in Universe is not a good idea(since the calculations will be done at Semantic layer instead of Database layer),as of now you can go for that as the variables created in Analytic/Calc Views are not exposed to reporting layer.

    Rgds,Murali

    Add comment
    10|10000 characters needed characters exceeded

    • Former Member

      Hi Zeeshan,

      there are some points to remember.

      1) How you are loading data into HANA from Source. If you are using realtime replication then i think you can not particular fields, you have to replicate entire table. If you are using BODS, then do the proper transformations and load the data into HANA.

      2) As you want to use Key Figures(Measures) you have to go for Analytic Views. If you create Analytic View, then you can select measures only from one table.

      3) If you want to do reporting on measures from different tables, then go for calculation views. One of the main objective of SAP HANA is "move the calculations to Data Base Layer". So try to avoid to do calculations at semantic layer and use the functionalities of SAP HANA.

      If you want to use Explorer, no need to create Universe, directly create Information View and use it.You can connect explorer using Universe also.

      Regards,

      Venkatesh

  • Mar 19, 2012 at 12:07 PM

    HI Zeeshan,

    1. Could you please let me know how you are loading data into SAP HANA?

    2. You need not to create any universe when you are doing reporting on explorer because explorer is directly connect to SAP HANA through JDBC drivers.

    3. You can use the Analytical View instead of Attribute view because in the analytical view you can use the transaction table If your record are frequently changed. (Note: you cannot use measures in the attribute view)

    Hope it is useful.

    Regards,

    Neha

    Add comment
    10|10000 characters needed characters exceeded

    • Former Member Former Member

      Zee, answer to your question for 20th March 2012:

      There are two ways of getting data into SAP HANA.

      1. SLT Replication for SAP ERP (real-time replication)

      2. SAP Data Services 4.0 for anything non-SAP based (near real time replication)

      SAP BO Data Services uses Sourced based Change Data Capture (CDC) technique for Delta Mechanism. DS offers CDC from RDBMS business suites for Oracle and Microsoft.

      The field called DI_OPERATION_TYPE generates any of the following values based on the source data nature

      I for Inserted record

      B for Before-image of an UPDATE (basically change on date or any a numeric point field)

      U for after-image

      D for Delete.

      You basically have to do following steps in SAP BO Data Services and SAP HANA for such replication -

      1. Create CDC enabled Data Source

      2. Import the Source table(s) to HANA

      3. Create a Datastore in HANA

      4. Create MetaData in HANA

      5. Create batch job in SAP BO DS.

      6. Create dataflow (by adding transformations) in SAP BO DS.

      7. Add the target table connections, map the delta fields accordingly to your target tables in HANA

      If your RDBMS is other than Oracle and Microsoft, then SAP BO DS supports Time-stamp based Delta mechanism. Please refer to SAP BO DS Guide for more information*

      ----------------------------------------------------------------------------------------------------------------------------------------------

      Now coming back to your next question -

      Why would SAP HANA have a PSA kind approach 😊  It is a purely an IN MEMORY DATABASE. It does provide you Analytical and Calculation views ( we call it Information models which are      calculation models Aka Column views, Aka HANA Cubes, Aka HANA views) which are basically not storing any data at all. These are definitions that are used for only calculations (Analytic Views and Calculation Views).

      For Master Data: you can opt for Analytical View otherwise known as Dimension model.

      You can build

      Hierarchies

      Measures (note: measures has to be defined only from single fact table)

      Attributes

      For Calculations, basically your HANA Cubes

      Multiple fact tables or multiple Analytical views. This generally uses complex calculations using L and R language scripts (SQL Scripting in HANA).

      ------------------------------------------------------------------------------------------------------------------------------------

      Now coming to your next question (When to Use Universe)

      Should I build my Universe of Information model or just on table (stored in HANA)

      Here are the PROs and CONs using both approaches -

      Universes on Information models in HANA aka Calculation Views:

      1. Information model in HANA are re-usable

      2. Same models can be analysed in Analysis or Explorer

      3. Calculation Views allows more complex levels of data manipulation than Universes

      CONs:

      1. SQL Constraints apply to Information models which might impact Query performance (in Explorer)

      ii) Universes on HANA tables

      PROs:

      1. Universes are tailored made for such ad-hoc analysis

      2. Query time, and user defined filters at Universe level can be push to database using HANA In memory

      3. Join on tables is faster than join on Information models.

      CONs:

      1. If you already have Information Models in HANA there is no bridge to re-use or re-model these in HANA Studio again.

      ---------------------------------------------------------------------------------------------------------------------------

      Next and last question, connecting to SAP BO Explorer:

      With the BO FP03, SAP still uses JDBC connectivity to SAP BO Explorer.

  • avatar image
    Former Member
    Jun 03, 2012 at 12:43 PM

    Getting back to the actual business scenario instead of the technology;

    The user would like to see the pipe shipments; right now the released extractors (https://service.sap.com/sap/support/notes/1558737)  by SAP do not include the OIGS TD shipments; there will have to be specific extractors for the information required;

    The user would like to see shipments which are off-loaded; i have designed a model for this; i will post a short article over it soon;

    Add comment
    10|10000 characters needed characters exceeded

  • avatar image
    Former Member
    Jun 22, 2012 at 06:02 AM

    Zeeshan,

    Presuming that you are on ERP 6.0 and SP 20 (min) there are several options..

    a) create a generic extractor for TD shipments and release it for extraction. (can include delta functionality if it is based off a table or a view)

    b) Create a custom extractor and utilize the program RODPS_OS_EXPOSE to release your own custom extractor. (FM, or Query)

    c) Add an entry into the ROOSATTR table manually,  of the delivered extractor and release and test the extractor although SAP does not support this extractor.

    Add comment
    10|10000 characters needed characters exceeded