Skip to Content

SAP Design Studio 1.5: parallel processing & backend sessions


I am using DS 1.5 SP02 and I created a Dashboard which uses many DSources causing issues in the backend server because of the number of RFC sessions running in the sametime (one per query).

I tried to fix that problem by assigning to each queries a "processing group" Under 'Data Binding" in the query Datasource properties

Let's assume I have 6 Datasources assigned to 2 different processing groups (3 queries assigned to PG1 and 3 queries assigned to PG2)

Each processing group should consume only ONE session in the backend server (BW transaction sm04).

I also disabled "Variable merge" by seting "Merge prompts" to false in the application properties.

When I execute the report through the LaunchPad I can see as many sessions (6) in BW transaction sm04 as I have queries (6).

This means that the grouping does not work and my problem is not solved as I am expecting only 2 sessions (one per processing group).

Any idea what I am missing in my configuration?

Thanks in advance for your help.

Add comment
10|10000 characters needed characters exceeded

  • Get RSS Feed

1 Answer

  • Oct 22, 2015 at 01:19 PM

    Hi Nicolas - have you seen the chart shown in Karol's document? Design Studio 1.5: View on Parallel Data Source Execution

    It depends on what your application is doing; could you check and report back? 

    Add comment
    10|10000 characters needed characters exceeded

    • Hi Nicolas,

      an application has always "main" session. And then one session per processing group.

      In your scenario, you should see 1 session for the app, and 2 more sessions (as you have 3 DS / processing group -> 2 groups).

      Of course, you see also the sessions from the designer. If you start the designer, there are also open session for this application in "design mode". This is probably why you see more than those 3.

      Check your sessions in SM05, open new application in a browser and check then the difference.

      Btw, as the main session is unused, it makes sense to put some data sources also into the main session (simply not assigning any processing group to eg. 2 data sources)