Skip to Content

In BPC 7.0 for NW data entered through Input sheet and various other sources are stored in BW cubes in the backend. These data gets stored in delta fashion which is a common feature of BW system.

That is, if 1000 is there in the system for any account say X and it is overwritten by 0, 0 doesn’t get saved in the system instead the delta value between 1000 and 0 is saved in the system. i.e. -1000.

This increases the number of data stored in the system instead of having 1 record it has 2 in the back end. When the number of records in the system increases the performance degrades.

So to avoid this BPC has optimization functionality which helps compressing the data in the cube and maintains the number of records in the backend.

The Standard process chain for Lite and Full Optimization involves only in compressing the data and remodeling the application (in case of Full Optimization). It does not eliminate the zeros after compression.

Presence of zeros which are not going to be of any use will also impact the system performance. These zeros after compression can be eliminated by updating the process chain corresponding to optimization.

Steps involved in updating the Process chain for zero elimination:

1. Select the process chain for optimization.

2. Double click the Collapse process and change it to edit mode and select “switch on zero elimination” check box.

3. Now save the collapse process and try executing the optimization process the compressed zero values will get eliminated.

Note:

1. Changing the Collapse in either of the process chain will affect the other also.

2. If you want zero elimination to happen for only one of the optimization proces follow these steps,

  • create a process chain by copying the existing standard chains for optimization,
  • Create a new process variant for collapse and update that.

Thus the optimization process chain can be used in optimizing the application with zero elimination.

To report this post you need to login first.

1 Comment

You must be Logged on to comment or reply to a post.

  1. Jens Koerner
    This blog describes all the steps to configure the zero elimination in the BPC optimize.

    However, sometimes if there are old compressed request that do have zeros in them the optimize with zero elimination does not delete these old zeros. This is a known BW problem and the program RSCDS_NULLELIM allows you to delete these old zeros that the optimize with zero elimination does not catch.
    Problem is that this program runs into a meta data validation error on the time dimension for BPC cubes.
    Note “1470209 – BW report RSCDS_NULLELIM on InfoCube without time dimension” addresses this issue.

    On a related topic, check out Note “1471019 – Adjust sequence of process type about Full/Light optimize” to ensure you cube statistics are really up to date. This ensures consistent performance.

    Jens Koerner

    (0) 

Leave a Reply