Announcement

Collapse
No announcement yet.

Performance indicators (Otimize the software)

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Performance indicators (Otimize the software)

    I am currently doing a master's dissertation on this software in order to optimize it based on performance indicators. For now, I've been gaining experience interacting with the mirth platform, and I realized that there are ways to improve the sending / receiving of the number of messages over a period of time. Can you tell me ways to do it through some examples?

    To test, I am using two channels, one that sends and one that receives. The sender fetches the messages from a local folder, which is constantly being filled with messages every 30 seconds. The receiver receives the messages in a local folder and in a mysql database, recording the time of sending each message.

    Regarding the performance indicators of this software, is there any other indicator (other than messages)?

  • #2
    Hi Bruno ,
    You have a lot of factors to consider, in a scenario like you described, the time of processing the message will depend in the configurations of the channels for example if it is set for Development it will take longer but if it is set to Disable it will increase the speed. Also you have to take in account the source types you are using which in this case i think both are using FTP readers, this will also impact the processing versus an HTTP Listener or simply a Channel Reader.
    I would start by measuring performance indicators according to real deployed integrations, for example, tipically in a Health Institution a Vendor A sends an HL7 v2 by TCP to a Vendor B. Assuming that both vendors use MirthConnect as their integration Engine, how much time does it take for either to send or process the event? Which events are being sent? Are there transformations that impact the performance? and so on ...

    I'm quite interested in your master's dissertation if the theme is performance indicators of the integration Engine. Give me a reach and i will try to give you my input.

    Comment


    • #3
      Originally posted by Turanga_Fry View Post
      Hi Bruno ,
      You have a lot of factors to consider, in a scenario like you described, the time of processing the message will depend in the configurations of the channels for example if it is set for Development it will take longer but if it is set to Disable it will increase the speed. Also you have to take in account the source types you are using which in this case i think both are using FTP readers, this will also impact the processing versus an HTTP Listener or simply a Channel Reader.
      I would start by measuring performance indicators according to real deployed integrations, for example, tipically in a Health Institution a Vendor A sends an HL7 v2 by TCP to a Vendor B. Assuming that both vendors use MirthConnect as their integration Engine, how much time does it take for either to send or process the event? Which events are being sent? Are there transformations that impact the performance? and so on ...

      I'm quite interested in your master's dissertation if the theme is performance indicators of the integration Engine. Give me a reach and i will try to give you my input.
      Thank you very much for your feedback and interest

      For now, I am testing the different types of sources: relational (mysql, postgrees, oracle) and non-relational databases (mongodb).
      For both sources, I run for 3 hours (messages every 30 seconds) and at the end I check the number of messages processed per second.
      Then I will start by checking the other sources.
      * The messages sent are hl7 v2.

      Another thing, do you think I should test in development, production, raw or metadata mode?


      Do you think so?

      Thanks a lot for the help
      Last edited by Bruno; 05-18-2020, 05:23 AM.

      Comment

      Working...
      X