Skip to Content
avatar image
Former Member

Consuming CDS views in SAP BW

Hi,

what is the best practice to consume a CDS view in SAP BW?

Example scenario - let's say we have an SRM system with embedded BW running on HANA. We have built some ABAP managed CDS Views on top of the SRM tables. Now we would like to reuse the results of those views in SAP BW - combine it with some other data loaded into BW from other sources, use the BW authorization concept etc.

Thanks,

Marek

Add comment
10|10000 characters needed characters exceeded

  • Get RSS Feed

1 Answer

  • Best Answer
    avatar image
    Former Member
    Oct 22, 2015 at 08:05 PM

    Hi, Marek.

    If CDS view have no input parameters (i.e. it is a common view inside DB, not a table function), you can import it into BW as OpenODS View based on the corresponding view inside HDB and use it for reporting. Then put it into Composite Provider in order to make transformations available (Composite Provider can be the source of a transformation, OpenODS View cannot). In such way you can consume also HANA information views.

    But keep in mind, that OpenODS view is like a proxy object to HDB (not an ABAP dictionary), so it doesn't know of changes in its source. You have to keep changes in sync manually.

    Another way is a Virtual Provider based on FM (inside FM you can use Dictionary-managed CDS) to work only with ABAP objects, but it looks to cumbersome and unnecessary.

    Regards,

    Andrey.

    Add comment
    10|10000 characters needed characters exceeded

    • Former Member Former Member

      Hi, Marek.

      Did you mean a "general" datasource trough DB Connect, as it was in old BW systems? If so, this is the way, but SAP offers more flexibility and ease of integration via OpenODS views or direct consuming of HANA Models through Composite Provider: you can include HANA Models directly into Composite Provider and mix them with BW objects like DSO (this ability I can see in BW SPS10). In this case you have direct access to your data without buffering (like PSA) or Application Server interaction (at infopackage execution) and use all the HANA speed, because data is processed at DB side and is always "in sync" (you have no unnecessary persistence).

      Also there's some additional approaches of BW modelling with SAP HANA in SAP's help portal: here and here as entry points.

      Keep in mind: new features are sometimes not so stable😉 I've faced some troubles related to newly introduced features because they are a subject to change and can be adjusted in future releases, and sometimes not so nicely tested. But if we use them we can help to improve them, because they are really nice.

      Regards,

      Andrey.