4 Replies Latest reply: Oct 20, 2011 7:54 AM by Frank Siciliani RSS

Data Extraction Times

Brian Murphy

Currently testing out the SAP Connector. Its taking a long time to extract data from BW cubes, in excess of 10 minutes to pull thousands of lines.


This seems extremely long as we have reports pulling data from sql that retreive > 5,000,000 lines in under ten minutes.


Is there a reason for this slow rate of extraction? Does anyone have any tips, etc that may improve performance?



  • Data Extraction Times
    Miguel Angel Baeyens

    Hi Brian,


    I'm experiencing similar load times extracting from tables and reports in SAP. I'd say that is part the driver (there are faster drivers for other database managers) but above all, it depends on SAP itself. I've seen it working and the interface responds more or less to the timing, meaning it is not faster than the load from QlikView.


    I'm not a SAP expert, though, so you let's see if there are some tweak you o your SAP people can do in the server, transports or configuration, but it doesn't sound extremely slower to me.


    Hope that helps.


    Miguel Angel Baeyens

    BI Consultant

    Comex Grupo Ibérica

  • Data Extraction Times
    Sunil Jain

    In case of cube . you have to extract fact and dimention table seperately and then join it in to qlikview. it is fastest method that I experienced.

    • Re: Data Extraction Times
      Matthias Steiling

      Ich bin bis 20.10.2011 abwesend


      Ich werde Ihre Nachricht nach meiner Rückkehr beantworten.

      Bei einer kurzfristigen Lösung bitte ich Sie die Mail an Herrn Kruschke

      (-132) oder an Herrn Pohlschmidt (-139) weiterzuleiten oder sie telefonisch

      zu kontaktieren.



      Hinweis: Dies ist eine automatische Antwort auf Ihre Nachricht  "[SAP] -

      Re: Data Extraction Times" gesendet am 19.10.2011 10:22:49.


      Diese ist die einzige Benachrichtigung, die Sie empfangen werden, während

      diese Person abwesend ist.

  • Data Extraction Times
    Frank Siciliani

    The other option is to not use OLAP connector and pull from the underlying tables/DSOs. This is what we do and for most of the data we are able to get very good performance.