General Technical Discussions for Service

Get Involved. Join the Conversation.

Topic

    Venkat Tanneru
    Tool to analyse OsVC REST API response time to external...
    Topic posted March 15, 2018 by Venkat TanneruBronze Medal: 1,250+ Points 
    99 Views, 5 Comments
    Title:
    Tool to analyse OsVC REST API response time to external application
    Content:

    Hi All,

    We are using external applications to GET and POST records into Service Cloud. We are observing major fluctuation in response time. Is there any tool to monitor these response time or is there any way to get notified when response time reaches some threshold time limit.

    Comment

     

    • Suresh Thirukoti

      POSTMAN tool can help I think...they have monitoring aspect....

      https://www.getpostman.com/docs/v6/postman/monitors/intro_monitors

      ~Suresh

    • Venkat Tanneru

      Thanks Suresh. We are looking if there any tools that Service Cloud offers :)

    • Suresh Thirukoti

      Not sure Venkat!! But we always rely on Postman or wireshark for evaluating Web services.....but never did monitoring

      ~Suresh

    • Luuk

      Oracle always points to Fiddler with these types of questions.
      Take a look at support answer: https://cx.rightnow.com/app/answers/detail/a_id/6775/

      It's about timeouts when using the API. There are some steps in there to capture traffic using Fiddler.

       

    • Rajan Davis

      I think there are too many factors to give you a clear indication of why there may small differences in response times, but I would recommend taking a look at the PAPI reports and/or copying them and extending them to include the counts of objects.

       

      While this won't tell you the time it takes to process a  GET or POST request, it should tell you the number of objects you are interacting with for a given call.

      If your calls are interacting with several thousand objects, then there may be latency.

       

      Other things to look out for/consider:

      1. Certain object properties will add size to the JSON data which can compound the response time (e.g. HTML solution for answers, threads for incidents).
      2. JSON serialization isn't cheap; check on how data is serialized/deserialized. Here are tips on improving this for .NET.
      3. The programming language making the calls can have some impact, but this is a very pedantic thing. This (at best) can shave off a few seconds here and there, but there can be tradeoffs like larger memory usage and lower ease of debugging.
      4. The internet connection speed of the computer using the external application can have an effect. I will generally see a HUGE speed improvement running a Ruby/Node/Python script on AWS using dedicated hardware versus using the same script on my computer; however, dedicated hardware is optimized for this, it is the expected behavior when running on dedicated hardware.
      5. Depending on how your application works, you might benefit using asynchronous HTTP requests; however, if your application works in such a way that you are relying on the HTTP request to return some data before you can have it do something, this won't help.