Clovertech

UPOC Protocol

› Clovertech Forums › Cloverleaf › UPOC Protocol

Tagged: 

  • Creator
    Topic
  • March 25, 2021 at 2:12 pm #118681
    Devika Agarwal
    Participant

      I have a Tcl script that scans for PDF files in the specified IB directory. When a PDF is available, it encrypts the PDF, saves to archive folder, transfers the source PDF to the specified OB directory and then deletes from the IB directory. The script by itself works but doesn’t work when I use within Fileset-local protocol. That might not be the best idea, so I have also tried UPOC protocol with Advanced Scheduling. The tps every 0,20,40 second. This works but I have two concerns:

      1. Would there be an impact on the performance if the UPOC thread is set to run on timer every 0, 20 and 40 second, all day? If we have 10 more of those threads, what would be the impact because I need a script for 10 separate IB/sources?
      2. How can I make this work with Alerts Configurator? I tried last received and last update but they won’t work with UPOC.

      Thanks,

      Devika

    • Creator
      Topic
    Viewing 1 reply thread
    • Author
      Replies
      • March 26, 2021 at 4:32 am #118682
        Robert Kersemakers
        Participant
          1. Performance is always a tricky subject. I all depends on the configuration of your server, the workload involved etc.
            Imho I don’t think letting Cloverleaf check a directory every 20 seconds will have a lot of impact, even if you have 20 of those threads. However if all these threads will receive 200 PDFs every 20 seconds you may run into troubles.
          2. If I read correctly your tcl script is not generating a message for Cloverleaf, only processing the PDFs. Hence Cloverleaf is not receiving messages and you can’t use an Alert to monitor this.
            What you could do is generate a dummy message in your UPOC tcl script (set mh [msgcreate “dummy message”]; lappend displist “CONTINUE $mh”) and insert hcitpsmsgkill into your Inbound TPS of that thread. This way a message is received and you can put an alert on it, but the message is killed after being received.

          Zuyderland Medisch Centrum; Heerlen/Sittard; The Netherlands

          • March 29, 2021 at 11:57 am #118688
            Devika Agarwal
            Participant

              The volume would depend on the source and some maybe pretty high volume (3K PDFs/day). Is there another protocol to use my script that might be better? If I do have to use the UPoC protocol, where do I add the script? The script doesn’t run if I add in the Read TPS or Write TPS.

            • March 31, 2021 at 3:58 am #118691
              Robert Kersemakers
              Participant

                There is no (file) protocol that will take the filename as a message. Could be a new feature. So you need to make this yourself.
                You should use the upoc protocol and insert your script in the Read TPS with a certain Read Interval (or Advanced Scheduling).

                Zuyderland Medisch Centrum; Heerlen/Sittard; The Netherlands

              • March 31, 2021 at 10:51 am #118693
                Devika Agarwal
                Participant

                  I added my script to the Read TPS with 5 seconds read interval and nothing happened. It works with Advanced Scheduling but I have to run at 0,20,40 seconds which is still fine as long as I can get the alerts working. I added the code to create the dummy message but I am not seeing the IB message count increase with or without hcitpsmsgkill.

                • March 31, 2021 at 4:12 pm #118697
                  Devika Agarwal
                  Participant

                    Never mind. It just worked finally. Been a long day. Appreciate your help!

                • March 29, 2021 at 3:14 am #118686
                  Ab Lugtenburg
                  Participant

                    Hi,

                    At the point of peformance i can tell you that if you encrypt (base64 in my case) and the file are big (say about 500 mb  or bigger) you have a performance issue under windows 2016. At our site all i made an extra site with these threads so the regular threads in the master site did not have a problem (my hl7 adt etc)

                     

                    • March 29, 2021 at 11:59 am #118689
                      Devika Agarwal
                      Participant

                        Which protocol you are using? Can you please share configuration details if you don’t mind? We do have a separate site where all these threads would be running.

                  • Author
                    Replies
                  Viewing 1 reply thread
                  • You must be logged in to reply to this topic.