Scenario for Using Write-Optimized DataStore Objects

A plausible scenario for write-optimized DataStore objects is exclusive saving of new, unique data records, for example in the posting process for documents in retail. In the example below, however, write-optimized DataStore objects are used as the EDW layer for saving data.

There are three main steps to the entire data process:

..

       1.     Loading the data into the BI system and storing it in the PSA

        At first, the data requested by the BI system is stored in the PSA. A 

        PSA is created for each DataSource and each source system. The PSA 

        is the storage location for incoming data in the BI system. Requested 

        data is saved, unchanged, to the source system.

       2.    Processing and storing the data in DataSource objects

       In the second step, the data is posted at the document level to a write-

       optimized DataStore object (“pass thru”). The data is posted from here 

       to another write-optimized DataStore object, known as the corporate 

       memory. The data is then distributed from the “pass thru“ to three 

       standard DataStore objects, one for each region in this example. The 

       data records are deleted after posting. 

       3.   Storing data in InfoCubes

       In the final step, the data is aggregated from the DataStore objects to 

       various InfoCubes depending on the purpose of the query, for example 

       for different distribution channels. Modeling the various partitions 

       individually means that they can be transformed, loaded and deleted 

       flexibly.