DMN Validation + Save taking too long

[Aman Agrawal]

Hi, I am using Business Central to design DMNs. However when making any change to a DMN, it is taking over 1.5 hours to validate and save one change. 

It takes very less time (~1 second) for files with less number of inputs and included models but I have a file with around 12 included models and this file is further included into another DMN file. So to make any modification and save this file it takes around 1.5 hours.

Issue seems to be usage of single CPU which is reaching 100%. Even if I run business central on a machine with 4 CPUs, it uses just 1 of it.

Please let me know if I am doing something incorrectly.

[Matteo Mortari]

Hi Aman,

That is a very strange behaviour indeed; the best suggestion here is to share a reproducer, so that the problem can be investigated on your use-case.

We had already benchmark on different parts of the code, for example even the decision table validation, so this seems to be some problem connected with the environment and/or specific use-case you have on hand. 

Without a reproducer it would be unfeasible to guess what's going on, so:

• I can suggest for you to try out if the problem is the same using the VSCode DMN Editor and modeling your DMN in your kjar project using that, or trying out with the Kie Sandbox if that helps. I'm positive other colleagues may have additional suggestions here wrt BC and/or BC alternatives.
  
  
• If the problem is the confidentiality of your reproducer, if you have a Red Hat Subscription, do raise a Customer Case and attach confidentially the reproducer there instead
  

Hope this helps!

MM

--
You received this message because you are subscribed to the Google Groups "Drools Usage" group.
To unsubscribe from this group and stop receiving emails from it, send an email to drools-usage...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/drools-usage/a0ef5f32-38a8-4a25-a6a4-f1e09d4399a6n%40googlegroups.com.

--

Matteo Mortari

linkedin.com/in/matteomortari

youtube.com/MatteoMortari

[Aman Agrawal]

Hi Matteo,

Thanks for the quick response.

I am afraid I won't be able to share the files as they are confidential however I can share a screen recording of a part of it. I can also provide all the metrics needed for debugging the issue. Aso, I will try out the suggestions given by you and see if it helps me. And no, I do not have a Red Hat Subscription.

Thanks and Regards,

Aman

To view this discussion on the web visit https://groups.google.com/d/msgid/drools-usage/CAHOoiEqq5iCwqiMwD0nQ0WG8opKz0s99zrHAdAmk_arOsFPb1A%40mail.gmail.com.

[Aman Agrawal]

Hi,

I am able to save files quickly using VS Code and the extension. However, when running 'mvn clean test', it is taking forever at this step "Analysing decision tables in DMN Model "[model_name_here]" for the file which had 12 models included.

Any option to skip it/make it faster and directly run tests?

[Matteo Mortari]

Hi Aman,

it would really help if you could share at least some other DMN, with a _similar_ table as the one contained in [model_name_here].

I can understand your original DMN model is confidential and you don't have a Red Hat subscription to share confidentially your actual DMN model. 

If you are able however to build a synthetic, similar, "example" DMN model, which also takes the same long time to build your kjar, we can take a look into that instead, to try to investigate the real root cause.

Let us know :)

At the same time, you can use this configuration to skip DT analysis while performing your KJAR build: 

https://docs.drools.org/7.73.0.Final/drools-docs/html_single/index.html#dmn-validation-con_dmn-models

Hope this helps!

MM

To view this discussion on the web visit https://groups.google.com/d/msgid/drools-usage/CABmktiL%2B_eYdmnb0qjhiU%3DrLfTisLtOECL26zmpA%3DhxmrVEbeA%40mail.gmail.com.

[Aman Agrawal]

Hi Matteo,

Sure, I can do that over the weekend. It will take some time to replicate the big DMN

Until then skipping DT Analysis has helped a lot. This has also fixed the issue for business central. ANALYZE_DECISION_TABLE - this was the time taking part.

Thanks a lot.

To view this discussion on the web visit https://groups.google.com/d/msgid/drools-usage/CAHOoiEoBMZEuqvqiZW2Uus3MyA-W4gs3MJL2CfX2nSGOtrBXGw%40mail.gmail.com.

[Matteo Mortari]

Hi Aman,

thank you for the confirmation; glad to hear it un-blocked on your way forward with you kjar project.

When you have the synthetic reproducer DMN example ready, if you prefer, you could also directly attach it while create a JIRA here: https://issues.redhat.com/projects/DROOLS

Or, even better, creating a kjar reproducer on github and making the link in the JIRA on this mailing-list

Hope this helps!

MM

To view this discussion on the web visit https://groups.google.com/d/msgid/drools-usage/CABmkti%2B4ijaNYAHqiROfGRsL58-BiXS7A3yroxy95tqAjyNVFA%40mail.gmail.com.

[Aman Agrawal]

Sure, I will do it. Thanks a lot.

To view this discussion on the web visit https://groups.google.com/d/msgid/drools-usage/CAHOoiErD03C%3DLx5hnzRuLfzg6FkfuXPwLS%3DvcS88cD7dkvON5A%40mail.gmail.com.

[Aman Agrawal]

Hi Again,

I am facing the same issue again. Even after disabling ANALYZE_DECISION_TABLE, it's still taking very long to save in Business Central. Any other solution?
Reason why I cannot use VS Code extension is - copy and paste text does not work in the extension.

Any help would be appreciated. Thanks & Regards.