Internal Table with 22 Million Records

Hello,

I am faced with the problem of working with an internal table which has 22 million records and it keeps growing. The following code has been written in an APD. I have tried every possible way to optimize the coding using Sorted/Hashed Tables but it ends in a dump as a result of insufficient memory.

Any tips on how I can optimize my coding? I have attached the Short-Dump.

Thanks,

SD

  DATA: ls_source TYPE y_source_fields,
        ls_target TYPE y_target_fields.

  DATA: it_source_tmp TYPE yt_source_fields,
        et_target_tmp TYPE yt_target_fields.

  TYPES: BEGIN OF IT_TAB1,
          BPARTNER TYPE /BI0/OIBPARTNER,
          DATEBIRTH TYPE /BI0/OIDATEBIRTH,
          ALTER TYPE /GKV/BW01_ALTER,
          ALTERSGRUPPE TYPE /GKV/BW01_ALTERGR,
          END OF IT_TAB1.

  DATA: IT_XX_TAB1 TYPE SORTED TABLE OF IT_TAB1
        WITH NON-UNIQUE KEY BPARTNER,
        WA_XX_TAB1 TYPE IT_TAB1.

  it_source_tmp[] = it_source[].

  SORT it_source_tmp BY /B99/S_BWPKKD ASCENDING.
  DELETE ADJACENT DUPLICATES FROM it_source_tmp
                        COMPARING /B99/S_BWPKKD.

  SELECT BPARTNER
          DATEBIRTH
    FROM /B99/ABW00GO0600
    INTO TABLE IT_XX_TAB1
    FOR ALL ENTRIES IN it_source_tmp
    WHERE BPARTNER = it_source_tmp-/B99/S_BWPKKD.

  LOOP AT it_source INTO ls_source.
    READ TABLE IT_XX_TAB1
      INTO WA_XX_TAB1
      WITH TABLE KEY BPARTNER = ls_source-/B99/S_BWPKKD.
    IF sy-subrc = 0.
      ls_target-DATEBIRTH = WA_XX_TAB1-DATEBIRTH.
    ENDIF.

    MOVE-CORRESPONDING ls_source TO ls_target.
    APPEND ls_target TO et_target.
    CLEAR ls_target.
  ENDLOOP.