4 Replies Latest reply: Mar 23, 2012 11:44 AM by Victor Angoa Camacho RSS

    memory dumps in SAP

    ioannis giakoumakis

      Hi everybody.

      We are quite new in SAP connector and are facing this problem :

       

      While trying to load data into QlikView from SAP BW Queries using the SAP Connector version 5.6, we get a lot of memory dumps in SAP affecting performance (the load returns millions of records, it is not a small database). We tried to go around the problem by using the D, S and O parameter in Pseudo MDX statements, but without any success. We then tried to partition the QlikView load, but this way we are not sure if the queries return all the data we need, because the smaller the partition, the less data is returned by the same query (up to a limit). Any suggestions?

        • memory dumps in SAP

          Hi giakoum,

           

          not really a solution, rather a workaround.

          Transfer all records from the query with the APD into a DSO and then grab the data from the DSO with the connector directly from the table /BIC/A<DSO-NAME>00

           

          Best regards

           

          Andreas

          • memory dumps in SAP

            Hi


            we have the following: a generator that we generate a QVD per year.

            In previous years it may take 40 min. in the current takes 5 min elapses and as the current year increase this time.


            an example load current year:


            LET Date = text (year (Today ()));

            LET Name = 'BSEG' & text (year (Today ()));

            $ (Name): / / Accounting Document Segment

            Load *

            [Campos]


            SQL Select * [Fields] *

              from BSEG

            WHERE GJAHR = $ (Date);


            Example years:


            LET Date = '2009 ';

            LET Name = 'BSEG' & '2009 ';

            $ (Name): / / Accounting Document Segment

            Load *

            [Campos]


            SQL Select * [Fields] *

              from BSEG

            WHERE GJAHR = $ (Date);


            regards