Skip to Content
Technical Articles
Author's profile photo Dejan Kostic

How to write PaPM data to remote SAP BW system


SAP Profitability and Performance Management (PaPM), provides capability of writing result data directly to SAP BW infoprovider using Writer function. However, this option is available only in case  PaPM and SAP BW are residing on same SAP system. It is not possible to use writer function to write result data to infoprovider on remote SAP BW system.


This document will describe one of the ways that can be used to write PaPM data to remote SAP BW system. Data flow will have to be manually created on SAP BW system and data will be pulled from PaPM table and loaded to ADSO. It will also describe how to automate the entire process using Remote Function Adapter function and trigger execution from PaPM system. Single RFA function will be used to trigger PaPM calculation and execute data pull from remote SAP BW system.


In this example, copy of standard sample content environment SXI – IT Cost Management is used. However, any executable function from any PaPM environment can be used for same purpose.

Goal is to write results of allocation function ALSBU – Allocate Services to Business Units, from PaPM system to ADSO ZALSBU on remote SAP BW system.

To make it simpler, ADSO ZALSBU is created using same fields like in ALSBU function.

Idea is to create a data source on remote SAP BW system to access result data of ALSBU function from PaPM system using SAP HANA Smart Data Access. Datasource will then be connected to ADSO and data loaded like shown in dataflow bellow. Process chain will be used to automate the load. Process chain  will be executed from RFA function on PaPM system.


In order to be able to follow the steps described in the next section, apart from PaPM, user needs to have fundamental understanding and at least basic knowledge of SAP BW and ABAP. All required ABAP code is included in blog.




Steps on SAP BW System

All bellow steps are performed using Modeling Tools for SAP BW for SAP HANA Studio or using transaction RSA1.

Create Data source

Datasource will be used to access data directly from result table of ALSBU allocation function. Settings should be maintained like shown below, using schema that PaPM uses for creating result tables, usually SAP<SID>. Generated DDIC table for function ALSBU should be selected. Table name can be found in modeling environment, locate the function in the function hierarchy and click on Function attributes button. In the dialog box, navigate to Runtime attributes tab to find the DDIC Table name.

Switch to the Fields tab, select FS_PER_PROC_ field and maintain selection properties as mentioned bellow. Result table of PaPM function could contain result sets for different Process id’s. During every function execution, data from result table are being refreshed. Data with same process id are deleted and new data inserted. Selection on process is is important to avoid duplication during data extraction. For every run we want do extract only data created for process id used in current run not all data from result table.


Create Transformation

Transformation connects datasource ZDSALSBU with ADSO ZALSBU.

Both objects use same field names and all transformation rules will be automatically created.


Create Data Transfer Process

Create Data Transfer Process to load data using proposed settings.

Select Filter Button and then select Create Routine to maintain filter routine for FS_PER_PROC_ID_ field like shown below. Routine will be used to restrict data extracted from result table and select only data for current process id. In case you are not using multiple process templates/instances in your PaPM environment this step can skipped.

Select button bellow to open routine editor.

Enter bellow code in editor screen of filter routine and Save.

*&  Include           RSBC_SEL_ROUTINE_TPL

program conversion_routine.
* Type pools used by conversion program
type-pools: rsarc, rsarr, rssm.
tables: rssdlrange.
* Global code used by conversion rules
*$*$ begin of global - insert your declaration only below this line  *-*
DATA: l_req_id      TYPE rsbc_requid,
      ls_rssdlrange TYPE rssdlrange.
*$*$ end of global - insert your declaration only before this line   *-*
* -------------------------------------------------------------------
*     Fieldname       = FS_PER_PROC_
*     data type       =
*     length          = 000000
* -------------------------------------------------------------------
form c_FS_PER_PROC_
  tables l_t_range structure rssdlrange
  using i_r_request type ref to IF_RSBK_REQUEST_ADMINTAB_VIEW
        i_fieldnm type RSFIELDNM
  changing p_subrc like sy-subrc.
*       Insert source code to current selection field
*$*$ begin of routine - insert your code only below this line        *-*


      i_r_request->get_requid( IMPORTING e_request_id = l_req_id ).

      DATA(lr_request) = cl_rsbk_request=>create_from_db( l_req_id ).
      DATA(l_logid) = lr_request->get_logid_pc( ).

      CHECK l_logid IS NOT INITIAL.

      SELECT SINGLE * FROM rspcrunvariables INTO @DATA(ls_var)
        WHERE meta_log = @l_logid AND fnam = 'PROC_ID'.

      ls_rssdlrange = VALUE #( fieldname = i_fieldnm option = 'EQ'
                                   sign = 'I' low = ls_var-low ).

          fieldname = i_fieldnm.
      IF sy-subrc = 0.
        <ls_range> = CORRESPONDING #( ls_rssdlrange ).
        APPEND ls_rssdlrange TO l_t_range.

    CATCH cx_root.

*$*$ end of routine - insert your code only before this line         *-*


Create Process Chain

Process chain contains only Start step and DTP execution. Depending on infoprovider, data activation step would be required in some cases. In case it’s planning ADSO, steps for switching real-time load behavior would be required as well.


Steps on PaPM System

Remote Function adapter function will be used to execute custom function module which will trigger process chain on SAP BW system, pull PaPM data and load it to ADSO. In this example RFA function uses allocation function as input, therefore, it will first execute allocation function, refresh data in result table and then execute Process chain on remote system. For this to work, it is required that allocation function ALSBU is set as executable. Allocation function should not be used as activity in any process template. In case ALSBU is used as an activity, it will have to be manually executed first, before executing RFA function.


Create Table Type

When External Function is used as Remote Function Adapter Type, it receives data from input function using input parameter IT_DATA. If you open rules tab, template will be created and you can see structure of IT_DATA. It basically corresponds to output table of input function ALSBU.

Therefore, you can use generated DDIC table for function ALSBU directly as a line type for your table type.


Create Function Module

Function module needs to have a signature like in RFA function template mentioned before and should be created with bellow source code. It receives IS_AL which will be used for getting process id selected during run. RFA function does not process any data, therefore, IT_DATA all though is necessary in function signature, will not be used in coding. Changing parameter CT_MSG table is used for passing messages to application log and it will propagate messages from process chain execution. Only thing that needs to be changed in coding is name of RFC destination and process chain that needs to be executed. Process ID will be taken from application log and passed as variable to Process Chain. Filter routine in DTP will get process id from process chain and use it to create a selection condition for data extraction.

FUNCTION zstartpc.
*"*"Local Interface:
  DATA: lv_chain     TYPE rspc_chain,
        lv_rfcdest   TYPE rfcdest,
        lv_logid     TYPE rspc_logid,
        lv_status    TYPE rspc_state,
        lt_variables TYPE rspc_t_variables,
        lt_msg       TYPE rspc_t_msg.

  lv_chain = ''.”PC that should be executed
  lv_rfcdest = ''.”RFC Destination
  lt_variables = VALUE #( ( fnam = 'PROC_ID' opt = 'I' sign = 'EQ' low = is_al-proc_id ) ).

      i_chain       = lv_chain
      i_synchronous = abap_true
      i_t_variables = lt_variables
      e_logid       = lv_logid
      failed        = 1
      OTHERS        = 2.
  IF sy-subrc <> 0.
* Implement suitable error handling here

      i_chain  = lv_chain
      i_logid  = lv_logid
      e_status = lv_status.

  CHECK lv_status CA 'RXJ'.

      i_chain     = lv_chain
      i_logid     = lv_logid
      i_dont_poll = abap_true
      e_t_log     = lt_msg.


  ct_msg = CORRESPONDING #( lt_msg ).



Creating Remote Function Adapter Function

Finally, RFA function can be added to the environment. This is the only change that needs to be done in existing PaPM environment. RFAWR function uses ALSBU as input function and calls the function module ZSTARTPC. Function itself does not do any data processing and result is not used, therefore, input data are not needed. To improve performances a bit, a condition that doesn’t select any data is added to the input tab.

Process chain of RFA function executes first ALSBU function and then RFAWR function itself. RFAWR function itself does not process any data but only triggers the process chain from remote SAP BW system to load ALSBU result data to ADSO.



Execution and Results

Once all previous steps are performed RFA function RFAWR can be executed.

After the RFAWR function is executed with process template SXI_P_, application log shows that function ALSBU created 126.515 records and Function RFAWR processed 0 records.


If you check ADSO  ZALSBU on SAP BW system, you will see that 126.515 records are loaded and DTP Selection FS_PER_PROC_ID_ = SXI_P_ was applied.


Hopefully, this blog gave you an idea how PaPM result data can be written to remote SAP BW system. PaPM doesn’t directly support it since it would require generation of SAP BW objects on remote system, however, it can still be achieved and automized with a bit of an extra effort from users side.

ABAP code can be reused in your examples and adjusted to support some additional requirements. Data flow and Process Chain on remote SAP BW system can be further enhanced and executed using same Function Module.




Assigned Tags

      You must be Logged on to comment or reply to a post.
      Author's profile photo Suvarna Ravindra
      Suvarna Ravindra

      Thank you for the detailed blog. I have a concern with regards to DDIC table name generated for allocation function. Your suggestion to specify DDIC table name of allocation function in datasource will not work when we move the changes to QA and production system since DDIC table name of allocation function will be different in QA and production system based on the client-id and the number range. Did you manage to handle this issue ?

      Author's profile photo Dejan Kostic
      Dejan Kostic
      Blog Post Author

      Hi Suvarna,


      You could for example create a custom table and use it in datasource instead of PaPM generated table. In your environment you would need to create model table function to connect to  your custom table and a writer function that uses allocation as input and model table as output.


      Best regards,


      Author's profile photo Denis Vikhrev
      Denis Vikhrev

      Hi Suvarna,

      All Executable and Model Table functions from the Environment are transports the fixed names of their Y-tables. Even if different clients and Range intervals are assigned to Dev-Qual-Prod systems, those Y-tables will have names from the origin system (Dev).

      However, some of BW Tech Attributes will be auto-generated with new names.

      Dejan Kostic, please correct me if I am wrong.

      Best regards,

      Author's profile photo Suvarna Ravindra
      Suvarna Ravindra

      Hi Denis,

      My PaPM system is on SP07 and system is generating Y-table with different name in QA and Production system for  all the functions including  model table and  executables.

      It would be nice to have fixed Y-table name for executables and model table function across system landscape. May I know the SP level of your system ?


      Ravindra Suvarna


      Author's profile photo Denis Vikhrev
      Denis Vikhrev

      Hi Suvarna,

      Are you using PaPM 3.0?

      I checked in different systems from SP06 to SP12 based on BW and S/4HANA systems, and it seems that Y-tables are always kept their names if you use standard transport functionality.

      However, I remember that for FS-PER 2.0 this cross-/single- client option was the Environment function setting in the Advanced tab. So in the old FS-PER version, it was possible to define system generation behavior for each specific environment.


      Best regards,