Uploaded image for project: 'Kogito'
  1. Kogito
  2. KOGITO-4257

Importing and modeling decision models is too slow for productive modeling

    XMLWordPrintable

    Details

    • [QE] How to address?:
      ---
    • [QE] Why QE missed?:
      ---

      Description

      Hi! We have been evaluating the use of Kogito in our projects for a while now and are excited about it, it's a great addition to the DMN ecosystem! However, while using it, we noticed that the import of decision models is becoming increasingly slower. It is now so slow that it limits productive modeling.

      Our largest decision models contain about 50 entities (decisions, business knowledge models, inputs), which should not be much for an extended decision model. Other modelers take a few seconds to load the decision models and then work without any problems. However, when using the current version of Kogito under Chrome, it takes over 1:30 minutes to import the decision models. Under Firefox, the import is not possible at all, because Kogito wants to allocate too much RAM. In addition, functions like "explore diagram" break the entire page regardless of the browser. The problem occurs with all users who have tested the decision models, so it does not depend on the browser or hardware used.

      Here are a few numbers: The latest version of Kogito (7.47.0.Final) needs 1:34 minutes and consumes 1.291 GB (! ) for a decision model with the size of 21 KB under Chrome. With an older version (7.42.0-SNAPSHOT) the model took 42 seconds and consumed 269 MB. Still a lot, but acceptable.

      Is this a known regression? I have attached an example model that produced the numbers mentioned. Thank you in advance!

        Attachments

          Activity

            People

            Assignee:
            karreiro Guilherme Gomes
            Reporter:
            jooas Jonas Tamimi (Inactive)
            Tester:
            Jozef Marko Jozef Marko
            Votes:
            2 Vote for this issue
            Watchers:
            6 Start watching this issue

              Dates

              Created:
              Updated:
              Resolved: