Skip to Content
avatar image
Former Member

Multiple sessions in R/3 generated by RFC adapter for the same message in PI during synchronous communication


During a synchronous call from FileàRFC in SAP R/3, the RFC throws some of the following locking errors:

E, 022, The group data for the material 00309157241 is locked by MDMREMOTE;

E-031-Characteristic Characteristic SHADECODE is being processed by MDM is bein;

The RFC is a wrapper RFC that internally makes nearly 10 calls to different RFCs based on conditions. This scenario was configured to eliminate ccBPM in PI within which 10 different calls to RFCs were handled synchronously. Since then, we have been facing more problems with data being locked or used or blocked by the same user that PI uses to log in to R/3 during an RFC call.

These locking errors are not thrown for all materials but observed for many. Some locking errors have been found even in the processing of other master data using different RFCs when the R/3 server is slow.

Message processing in PI was monitored for such cases. It was found that, during the synchronous call to RFC in SAP R/3 by PI, if the RFC does not complete/return something in a certain amount of time (2 minutes observed in the audit log of file sender communication channel), PI opens up another session in R/3 for the same material. Either of these sessions now face the problem of objects being locked, that too by the same user and for the same material. Upto 8 sessions have been seen generated by PI for the same message. The parameters set in the communication channels were checked but did not correspond to the above.

We have been facing this problem in our Production system and it has now become a major issue.

Can someone help to understand as to why this could happen? Is it possible that some parameters or attributes in the PI server or java engine or the RFC adapter module could be triggering another session?


Priyanka S

Add comment
10|10000 characters needed characters exceeded

  • Follow
  • Get RSS Feed

1 Answer

  • avatar image
    Former Member
    Nov 21, 2012 at 10:18 AM

    does RFCs you are using having table parameter, if you can submit in array this situation can be avoided,

    how you have made your file to RFC synchronous (have you used bean/adapter module?)

    another option is store all file records in custom table and then execute this RFC in one by one call after getting commit work for each record.

    Add comment
    10|10000 characters needed characters exceeded

    • Former Member Former Member


      When PI sends a sync. message to RFC, it is taking longer to respond. Since it takes longer, PI creates another session in R/3 for the same message in 2 minutes. Since both these sessions are active for the same materials, either of them face a locking/blocking issue.

      I changed the synchTimeout parameter of CallSapAdapter to 3 minutes in the module configurations of the File sender communication channel. It still created a new session after 2 minutes.

      Can you please guide where in NWA does this parameter needs to be changed?