BAPI_GOODSMVT_CREATE in parallel mode causes "plant data of the material XXX is locked" error

Question

Currently, We're developing mass GI posting by parallelism, to reduce work time dramatically, because we have to post GI item about 300k+ at one day.

Problem

I know there is a restraint about BAPI_GOODSMVT_CREATE - You can't post same material simutaneously because of table lock, so I grouped materials like this :

It seems reasonable, so I make parallel job for each material number, which total 7 in example.

When I try process this process, SAP refused to process with error :

M3 897 : The plant data of the material XXX is locked by the user XXX

Question

The error is about plant data, not stock data - so it's really confusing about why this error happening. Is there a more restraint about BAPI_GOODSMVT_CREATE?

Note 369518 won't help about this situation.

Any help would be appreciated!

Program Code

LOOP AT gt_matnr_collect INTO wa_matnr_collect.

    WAIT UNTIL g_progs <= pa_wpnum.

    g_progs = g_progs + 1.

    CALL FUNCTION 'Z_PP_STAMP_PARA_DISTRIBUTE'
      STARTING NEW TASK g_task DESTINATION IN GROUP DEFAULT
      PERFORMING return_stamp_para ON END OF TASK
      EXPORTING
        p_matnr = wa_matnr_collect-gi_code
        p_spmon = p_spmon
      TABLES
        p_list  = gt_main_para_list
        p_item  = gt_main.

    g_task = g_task + 1.
    g_sprog = g_sprog + 1.
  ENDLOOP.

gt_matnr_collect looks like this :

gt_main_para_list looks like this :

gt_main looks like this :

FM 'Z_PP_STAMP_PARA_DISTRIBUTE'

FUNCTION z_pp_stamp_para_distribute.
*"----------------------------------------------------------------------
*"*"Local interface:
*"  IMPORTING
*"     VALUE(P_MATNR) TYPE  MATNR_D
*"     VALUE(P_SPMON) TYPE  SPMON
*"  EXPORTING
*"     VALUE(E_MATNR) TYPE  MATNR_D
*"  TABLES
*"      P_LIST STRUCTURE  ZSPP6030_PA_LIST
*"      P_ITEM STRUCTURE  ZSPP6030_PA_ITEM
*"      R_RESULT STRUCTURE  BAPIRET2 OPTIONAL
*"----------------------------------------------------------------------


  DATA : gs_head      LIKE bapi2017_gm_head_01,
         gs_gmcd      LIKE bapi2017_gm_code,
         gs_doc       LIKE bapi2017_gm_head_ret,
         gt_item      LIKE TABLE OF bapi2017_gm_item_create
                                  WITH HEADER LINE,
         gt_return    LIKE TABLE OF bapiret2 WITH HEADER LINE,
         gs_return    LIKE bapiret2,
         group_id(22) TYPE c,
         l_begda      LIKE p0001-begda,
         g_last_date  TYPE d,

         lv_matnr     TYPE ztpp5360-gi_code,
         lv_batch     TYPE ztpp5360-gi_batch,
         lv_aufnr     TYPE ztpp5360-arbpl,
         lv_cnt(3)   TYPE c.

  CONCATENATE p_spmon '01' INTO l_begda.

  e_matnr = p_matnr.

  CALL FUNCTION 'RP_LAST_DAY_OF_MONTHS'
    EXPORTING
      day_in            = l_begda
    IMPORTING
      last_day_of_month = g_last_date
    EXCEPTIONS
      day_in_no_date    = 1
      OTHERS            = 2.

  CONCATENATE p_matnr '*' INTO group_id.

  LOOP AT p_list WHERE group_id CP group_id.

    CLEAR : lv_matnr, lv_batch, lv_aufnr, lv_cnt.
    SPLIT p_list-group_id AT '-' INTO lv_matnr lv_batch lv_aufnr lv_cnt.

    CASE lv_batch.
      WHEN 'EMPTY'.
        lv_batch = ''.
    ENDCASE.


        p_list-status = 'WIP'.
        MODIFY p_list.

        LOOP AT p_item WHERE group_id = p_list-group_id.

          MOVE : p_item-gi_m_q   TO gt_item-entry_qnt,  
*             'KG'             TO gt_item-entry_uom,  
                 'X'              TO gt_item-withdrawn,  
                 '3100'           TO gt_item-stge_loc.

          MOVE : '261'            TO gt_item-move_type,  
                 p_item-gi_code  TO gt_item-material,   
                 p_item-werks    TO gt_item-plant,      
                 p_item-aufnr    TO gt_item-orderid,    
                 lv_batch        TO gt_item-batch.           

          APPEND : gt_item.  CLEAR : gt_item.
        ENDLOOP.

        MOVE : g_last_date    TO gs_head-doc_date,      
           g_last_date    TO gs_head-pstng_date,    
           '03'           TO gs_gmcd-gm_code.

        REFRESH : gt_return.
        CLEAR   : gt_return, gs_doc.

        CALL FUNCTION 'BAPI_GOODSMVT_CREATE'
          EXPORTING
            goodsmvt_header  = gs_head
            goodsmvt_code    = gs_gmcd
          IMPORTING
            goodsmvt_headret = gs_doc
          TABLES
            goodsmvt_item    = gt_item
            return           = gt_return.

        READ TABLE gt_return WITH KEY type = 'E'.

        IF sy-subrc = 0 OR gs_doc IS INITIAL.
          CALL FUNCTION 'BAPI_TRANSACTION_ROLLBACK'.

          LOOP AT p_item WHERE group_id = p_list-group_id.
            p_item-ecode = 'E'.
            p_item-msgtx = gt_return-message.
            p_item-aedat = sy-datum.
            p_item-aezet = sy-uzeit.
            p_item-aenam = sy-uname.

            MODIFY p_item.
          ENDLOOP.

          p_list-status = 'ERR'.
          p_list-msgty = 'E'.

          CONCATENATE gt_return-message '/' gt_return-id '/' gt_return-number INTO p_list-msgtx.

          MODIFY p_list.
        ELSE.
          CLEAR : gs_return.
          CALL FUNCTION 'BAPI_TRANSACTION_COMMIT'
            EXPORTING
              wait   = 'X'
            IMPORTING
              return = gs_return.

          IF gs_return-type = 'E'.

            p_list-status = 'ERR'.
            p_list-msgty = 'E'.
            p_list-msgtx = gs_return-message.

          ELSE.

            LOOP AT p_item WHERE group_id = p_list-group_id.
              p_item-gi_m_mblnr = gs_doc-mat_doc.
              p_item-gi_m_mjahr = gs_doc-doc_year.
              p_item-ecode = 'S'.
              p_item-msgtx = ''.
              p_item-aedat = sy-datum.
              p_item-aezet = sy-uzeit.
              p_item-aenam = sy-uname.

              MODIFY p_item.
            ENDLOOP.

            p_list-status = 'FIN'.
            p_list-msgty = 'S'.
            p_list-msgtx = ''.

            MODIFY p_list.
          ENDIF.
        ENDIF.

        REFRESH : gt_item.
        CLEAR : gs_head, gs_gmcd.

        WAIT UP TO 2 SECONDS.
      ENDLOOP.
    ENDFUNCTION.

Have you done any analysis on where the lock is being applied from? That's the first thing I'd look into. — Skin, Apr 14 '22 at 00:58
Can't reproduce problem, because I've already done previous month process. I suspect not only material number, but also production order can be restraint at the same time - so I can't do GI posting with same material number OR same production order at the same time. — Park Chan Seul, Apr 14 '22 at 01:15
You said your parallel job is set to run by material number, and you've said 7 jobs. Does that mean the first job for material A runs 5 parallel jobs in background as well or does it run them synchronously/sequentially? I would think the latter approach would work. — Skin, Apr 14 '22 at 01:29
It's working like : Create aRFC job for 7 material code simutaneously, which are doing BAPI, which have line item which group by material code. 1st aRFC job : Doing A Material Code GI job posting - 5 line items 2nd aRFC job ; Doing B Material Code GI job postiing - 1 line items .. .. — Park Chan Seul, Apr 14 '22 at 02:00
Hmmmm, interesting. I have to admit, without building your scenario myself and without knowing why, it’s quite tricky to diagnose. I would expect it to work but you’ve proven it doesn’t. — Skin, Apr 14 '22 at 02:45
Yes, It's kinda difficult situation - I can post fragment of my program if you'd like to see. — Park Chan Seul, Apr 14 '22 at 04:22
Have you tried using a "where-used-list" on this message M3 897 in transaction SE91? If not you could put external breakpoints on all of the spots found there and then simply debug it (assuming you work with a small list of materials initially, if not then debugger sessions start popping up till the maximum number of available sessions is reached). — Jagger, Apr 18 '22 at 07:28
@Jagger Sadly, I can't debug this one right now, because it doesn't seem to be problem in Testing system, it happens only Production System, which I can use it again at the next month's first working day(we're using this process as backflush - only as CBO). But your advice seems to be helpful, I assume 2 standard function would be a problem - I'll external debugging at that point next time. — Park Chan Seul, Apr 18 '22 at 08:09

score -1 · Answer 1 · answered Mar 28 '23 at 13:47

-1

i dont know if this already solved or not. but if still not then.... i dont know about your case. but can you try to discuss with functional team. this is standard error that can be "pass" by configuration there is MM configuration about this one Locking activity (exclusive/late/shared). usually this will be EXCLUSIVE. you need to change it to LATE. try to discuss and explore this locking material config with your team

answered Mar 28 '23 at 13:47

user21510528

1

1

This answer would be more useful if it would explain where and how this configuration setting can be changed. – Philipp Mar 29 '23 at 10:00

BAPI_GOODSMVT_CREATE in parallel mode causes "plant data of the material XXX is locked" error

1 Answers1

Linked