ProcessFlow Server SQL configuration

Sort:
You are not authorized to post a reply.
Author
Messages
jamesraceson
Veteran Member
Posts: 52
Veteran Member

    I’ve been successfully working with ProcessFlow for many years; arguably one of the best interface/processing tools I have ever used.  Recently we have been trying to use the SQL Query node to grab data from other database systems (other than our Infor Lawson one).  Once you get the correct “lock and key” set up between drivers, permissions, and server access it works great.  Currently we have been using the SQL node with JDBC thin client for only Oracle databases.  We now want to use the SQL node for SQL databases.  I downloaded the correct SQL jar files and got this to work locally from my desktop.  When we went to test this on our hosted development server, the only way to get it to work by “running from server” was to only have SQL jar files in the applicable server folder (had to remove the Oracle jar files.  I was told by the hosting vendor that we had to choose either having Oracle or SQL JDBC, not both.  The ProcessFlow server could only be configured to look at one or the other JDBC files.  It works locally through ProcessFlow running both, but not at the server end.

     

    So I’m asking the residential gurus…is this true?  Can the ProcessFlow server only be configured for either SQL or Oracle JDBC and not both at the same time?  We are using ProcessFlow Integrator 9.0.1 (with the Eclipse version designer) but will be taking the plunge to IPA first quarter 2016.

    jdever
    Advanced Member
    Posts: 29
    Advanced Member

      Not true.

      I have created flows that access both MS-SQL & Oracle for the same client.

      You will need to "override" the configuration at the SQL node to point to MS-SQL (or Oracle).

      --

      Your hosting vendor may be stating that you can only have 1 "overall" configuration.

      In IPA 10 this has changed to allow multiple "overall" configurations.

      jamesraceson
      Veteran Member
      Posts: 52
      Veteran Member

        jdever,

         

        I'm using the "Overide JDBC" option for both my SQL and Oracle JDBC flows in ProcessFlow designer, but it still won't seem to work on the server side.  It works great locally as I can connect and retrieve information from both databases, just not when it is "run from server" (either through ProcessFlow designer or through PF Administrator by manual trigger).  Is there another "Override JDBC" option on the server side that you are referring to?

        jdever
        Advanced Member
        Posts: 29
        Advanced Member

          Make sure you have the proper JDBC jar files for both MS-SQL & Oracle in the $GENDIR/bpm/jar directory.

          jamesraceson
          Veteran Member
          Posts: 52
          Veteran Member

            jdever,

            The ones I use locally work perfect, so I'll have to verify with the hosting company they are using the same ones.  I'll contact them and have them install the ones that I'm using, retest, and update the post.  Thank you for the advice. 

            jamesraceson
            Veteran Member
            Posts: 52
            Veteran Member

              jdever,

              I tested again and to no avail, I could not get this to work.  Can you please let me how you are set up with your server settings that allows SQL JDBC to both SQL and Oracle databases?  This way I can take any configuration information that you have and replicate the same set up for our hosted environment and hopefully get this to work.

              jamesraceson
              Veteran Member
              Posts: 52
              Veteran Member
                I look forward to hearing back from jdever in hopes that there is some setup information that I'm missing.

                In the meantime, has anyone else been able to make this work successfully on the ProcessFlow server side (able to query both Oracle and SQL databases from the same server using JDBC in the SQL Query node)? I can get one or the other to work, but not both at the same time from the same server (even with the "Override" options). Any help would be appreciated.
                jdever
                Advanced Member
                Posts: 29
                Advanced Member
                  The UNIX server that is running pflow version 9 has the following directory:
                  $GENDR/bpm/jar with the following files:
                  ojdbc6.jar & sqljdbc4.jar.
                  --
                  I know that MS-SQL only wants 1 MS sql jar file inf the directory. i.e. Do not add sqljdbc.jar (for earlier versions 0f MSSQL).
                  I am not sure about Oracle. I have seen multiple ojdbc#.jar files in this directory with no problems.
                  I am attaching 3 print screens from the flow "badges.xml" that shows a MS-SQL query that updates an Oracle database with the badge number from the MS-SQL query.
                  jdever
                  Advanced Member
                  Posts: 29
                  Advanced Member
                    print screens of flow attached.
                    Attachments
                    jamesraceson
                    Veteran Member
                    Posts: 52
                    Veteran Member
                      jdever,
                      Thank you for the excellent information and follow up. Armed with these, I'm going back to our hosting vendor and see if we can make this work. I'll update this thread with my results. Either way it works out, thank you for your help with this. I appreciate it.
                      jamesraceson
                      Veteran Member
                      Posts: 52
                      Veteran Member
                        jdever,
                        Let the clouds part and music rain down!!! This worked perfectly! The key was to only have one SQL jar file in the jar directory. Oracle can have multiple jar files in the jar directory with no problem. Oddly enough, the main reason I needed this to work was because of the interfaces to our badge system as well. Thank you once again jdever with your help on this; greatly appreciated. As you know, with the limited resources available from Infor and most everything we do in ProcessFlow is "beyond the scope", it's help like this that makes it possible work out solutions to our problems. Thank you.
                        You are not authorized to post a reply.