3 Replies Latest reply: Jun 15, 2011 11:17 AM by Patrick Laredo RSS

    Encryption of Data

    Byron Van Wyk

      Hi Guys,

       

      I have a bit of a problem. I want to upload a document cause I want some assistance with removing loops and synthetic keys. Unfortunately the information is of a sensitive type. How can I keep my load script the same and the data behind it without anyone else seeing the actual information

       

       

      Cheers,

      Byron

        • Encryption of Data
          Patrick Laredo

          hi,

           

          go to "document properties", there is  a tab called Scrambling. Select the fields whose content you wish to scramble.  Make a copy of your doc beforehand!!!

           

          be selective on what you scramble, for example if your questions are about dates don't scramble your date fields.

            • Encryption of Data
              Byron Van Wyk

              How secure is this method Pat? Does QV claim or make mention in any documentation about the level of security that this provides a QV document.

               

              I am very conscious about protecting company information

               

               

              Thanks,

              Byron

                • Encryption of Data
                  Patrick Laredo

                  hi,

                   

                  I'm not a qv employee and I have no idea what they claim for this.

                   

                  This is what help says:

                   

                  Press this button to perform a random scrambling of the contents of selected field(s). Numbers are scrambled to numbers and text to text (spaces are kept). Once scrambled, the data cannot be recreated in its original form by QlikTech or anyone else. If you re-execute the script, however, the scrambling will be lost.

                   

                  However real life models can be daunting when uploaded as I doubt many people on this forum have the time it takes to deconstruct and then reconstruct entire applications. You are, in my opinion, more likely to get an answer if you cna break your problem down into chunks and load easily assimilated test datasets in a qvw.

                   

                  There are plenty of very good consultants out there who will come in and do a job for you which might be the case when looking at entire models. Here you will get answers that people already know or spring to mind quickly. It obvioulsy gets harder the more complex your problem.

                   

                  anyway that's my 2cts.