avatar image
Former Member

SAP HANA optimized Info Cube; Should we or Should we NOT Optimize and why?

Our current system is SAP BW (7.4 sp15) on HANA. We recently migrated to SAP HANA (2 weeks ago).

We have a customized Purchasing Order Cube which is currently a "Standard InfoCube". I have understood how it's structure (from E and F tables to only F tables) changes when we converts it into "HANA Optimized InfoCube", but one thing I am unable to understand how does it become helpful using from HANA?

I have 5 questions or confusions:

  1. Should we keep customizing the existing InfoCube in SAP BW, or should we move it into HANA by creating Advance DSOs, and building HANA views on top of those to have less or No dependability on SAP BW going forward.
  2. How converting into HANA Optimized Info Cube would help us. Should we or should we not convert it or any other ones and why?
  3. Once we convert it what would be next thing I have to do?
  4. Will all the existing BeX reports and OLAP Analysis will be fine or I have to change something there as well?
  5. At this moment we have to do few enhancements to existing PO Cube, so should we do those enhancement now then Optimize, or we can do it in Optimized once as well?

Please help me and explain these questions? I am very confuse and I have to present my suggestions to management soon.

Add comment
10|10000 characters needed characters exceeded

  • Get RSS Feed

3 Answers

  • avatar image
    Former Member
    Apr 03, 2017 at 08:39 AM

    Hi Tariq,

    1. aDSO is the only persistent layer that is going forward to be available as per SAP Roadmap (from SAP BW/4HANA). And adding to it aDSO is SAP HANA optimized object which will give better performance in HANA. So it would be better if you migrate to aDSO instead of cube.


    also you need to check below link for limitations.


    Known limitation for SAP BW 7.4(SP 15)

    • write-APIs ("direct-update" DSO)
    • Process chain variant:Request deletion in ChangeLog
    • Dynamic Tiering compatible (ABAP-based) activation/rollback
    • NLS connection for Advanced DSOs
    • Planning on Advanced DSOs
    • SPO using Advanced DSOs

    2. Above answer is applicable for this question also.

    3. After converting it to aDSO - you need to add transformation and mapping to new info provider and validate data

    4. You need to create new Bex Queries on top of new infoprovider and validate it with classical infoprovider queries

    Actually this is an upgrade project, you need to have the inventory list of cubes, classical DSO and consider LSA++ architecture to reduce data footprint as well.

    Add comment
    10|10000 characters needed characters exceeded

  • Apr 03, 2017 at 08:29 PM

    Hello Tariq,

    the general recommendation is to HANA-optimize all InfoCubes. Please use the SAP BW Checklist Tool (SAP Note 1729988) to see if there would be any limitations for your particular system ("Data Warehousing" in tool).

    The change is transparent to application and the analytic engine. No further adjustments necessary. Benefit is better loading and query performance.

    The only exception would be if you have custom code that accesses the InfoCube fact and/or dimension tables directly. But that's "worst practice" and should be changed to proper APIs anyway.



    Product Management SAP HANA DW

    Add comment
    10|10000 characters needed characters exceeded

  • avatar image
    Former Member
    Apr 10, 2017 at 01:52 AM

    Thank you very much Soundravel and Marc for your replies. I really appreciate Soundravel for detailed answer and covering my each question with separate answer. I am not sure if it would be as simple, but I am looking into it.

    Thank you one more time for both of you.


    Add comment
    10|10000 characters needed characters exceeded

Skip to Content