Large size HL7 messages on the engine.

Clovertech Forums Cloverleaf Large size HL7 messages on the engine.

  • Creator
    Topic
  • #119277
    Krishna Kumar
    Participant

      We are running Cloverleaf version 20.1 on Windows server. Our EMR system is EPIC.

      We are running an interface sending MDM messages to a downstream  Mosaiq. Recently we noticed that Epic started sending  larges messages of size more than 4 MB occasionally.  Down stream vendor Mosaiq is chocking on this since they can hold messages of size 4 MB or less. When every time this happen we have to manually remove the message from Cloverleaf queue to keep interface running.

      Question: Do any one else is facing a situation like this? If so how that is being fixed. Thanks in advance.

       

    Viewing 2 reply threads
    • Author
      Replies
      • #119278
        Jerry Sawa
        Participant

          This won’t help you much, but it will save you time manually removing the msgs from the Recovery Database until you find a solution.  In the Pre-Proc, check the size of the message and block it from going to Mosaiq if it is over 4mb.  You can add an entry to the log and also send an email notifying you msg was “killed”.

          This is some code I put in a pre-proc.  It doesn’t kill the msg, but it adds an entry in the log.

          set msg_length [ string length $msg ]
          ### Add log entry if msg is >= 10 MB ###
          if { $msg_length >= 10000000 } {
          echo “####### $tcl_script – Source = <$source_thread> — MSH.10 = <$msh_10>          — Length of msg = <$msg_length> — Message is *GE 10 MB”
          } else {
          echo “####### $tcl_script – Source = <$source_thread> — MSH.10 = <$msh_10>          — Length of msg = <$msg_length>”
          }

          • #121160

            Hi Jerry

            Can you share the tcl proc you are using for checking the file size .

             

            Thanks

            Raj

             

             

        • #119279
          Krishna Kumar
          Participant

            Thanks Jerry for your suggestion

          • #119280
            Jay Lontoc
            Participant

              I’m assuming the MDM from Epic contains embedded base 64 PDF. We had a similar issue with a downstream system and we ended up sending the Text version instead of the PDF. Another solution could be to send the PDF to your DMS application and Mosaiq can access the PDF there.

              You can also edit your ACK TPS to kill the inbound reply if you’re getting NAKs from Mosaiq. That way, the interface keeps running.

          Viewing 2 reply threads
          • You must be logged in to reply to this topic.