Info Kogito about: Apache incubator - production mode - modeller bpmn2.0 web based

[Antonio Delli Carpini]

Hi everyone,

I would like to find answers to the following questions.

1. Kogito is an Apache incubator project. Apache incubator projects become graduate projects after about 2 or 3 years.  Are there plans to promote a graduate projects?
2. Is there a project template for production mode? If so, where can I find it?
3. This article "https://medium.com/kie-foundation/kogito-online-dmn-bpmn-editor-and-chrome-extension-improvements-fe09353affc7" discusses an online (web-based) tool for modeling BPMN 2.0 flows. Where can I find this online tool today? Is there a reference?

Regards

[Alex Porcelli]

Graduation from incubation is not strictly tied to software maturity
or time. It depends on meeting the Apache Software Foundation's
policies and process requirements. This thread [1] outlines the
progress we have made and the challenges we are still working through.
I expect it will take about another year (hopefully less) to complete
the process.

Production usage is your decision. This is a mature technology that is
already being commercialized in different forms.

You can explore the online version here [2].

[1] https://lists.apache.org/thread/gmstcldbl65dy5nrjtf1omblrwryh8y2
[2] https://sandbox.kie.org/#/

-
Alex

> --
> You received this message because you are subscribed to the Google Groups "Kogito development mailing list" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to kogito-developm...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/kogito-development/6d07b8ff-2ab6-4238-ad1c-46743e9f5989n%40googlegroups.com.

[Antonio Delli Carpini]

Thank you, Alex.

So far I've used jbpm. I've been trying to follow the kogito project for about a year. I wanted to get some advice for minimal architecture in production mode with Kogito.

I would also like to understand if there are blocks in the bpmn modeler already prepared for SOAP invocations and REST invocations (GET/POST/PUT/DELETE/).

Regards.

[Antonio Delli Carpini]

Hi Alex,

I have more questions to ask:

1. Is there a way to use the sandbox instead of the vscode plugin?
2. If so, will the flows be automatically saved in the appropriate resources folder of the "running" module? Or will they be saved in specific database tables?
3. Is there source code for the sandbox in this repo "https://github.com/apache/incubator-kie-tools/"?
4. Are there any guides to understand how to configure the sandbox?

Regards

Il giorno mercoledì 23 luglio 2025 alle 15:15:38 UTC+2 Alex Porcelli ha scritto:

[Alex Porcelli]

Sandbox largely runs on browser only, it stores data in browser local
storage. It's up to the user if they want to store it in a git
repository to do so by connecting their account to sandbox and push
whenever they see the need.

Sandbox codebase is here:
https://github.com/apache/incubator-kie-tools/tree/main/packages/online-editor

You can learn more how to run Sandbox in this old post,
https://medium.com/@caponetto/deploy-your-kie-sandbox-to-openshift-f0b8442d046d

You have updated images in docker hub.


On Wed, Aug 13, 2025 at 10:40 AM Antonio Delli Carpini

> To view this discussion visit https://groups.google.com/d/msgid/kogito-development/9297356c-b2c9-4a02-b454-83532db97341n%40googlegroups.com.

[Antonio Delli Carpini]

Hi Alex,

is it possible to have multiple runtime nodes in the same kogito instance? (See attachment)

If so, will there have to be a particular configuration?

Are there any contraindications to having multiple replicas of the kogito-runtime module?

Regards

[Alex Porcelli]

This image seems related to distributed deployment; which I wouldn’t recommend. We experienced in past data loss using this setup (and you need add Kafka and JobService in the mix as well).

Please check the compact architecture, as it’s a reliable deployment that has been properly hardened. However, the compact architecture you can deploy as many container instance you want, but given some limitation with community job scheduler you’ll end up with only one instance being really active.

Regards,
_____________
Alex Porcelli
http://porcelli.me

To view this discussion visit https://groups.google.com/d/msgid/kogito-development/83255a07-25ff-485e-9c0b-8475d9e806fen%40googlegroups.com.

[Antonio Delli Carpini]

Hi Alex,

so if I understand correctly, even using the compact architecture the bpm engine cannot have other replicas because only one would work.

I would like to make a small observation with a use case.

Hypothesis: use compact architecture recommended by you.
Use case: there are 3 containers of the runtime module (A, B and C).
At time T0 only the runtime present on node A works, but nodes B and C are still active.
At time T1 node A stops working, then one of nodes B and C automatically
Does the engine automatically resume correct operation?
Or, to resume the correct functioning of the engine, must all the containers and the engine be restarted?
If so, I understand that the community version, even the compact architecture you recommended, no longer allows engine replication.

So the bpm engine can't have other containers and so it's a big kogito limitation.

Regards

[Alex Porcelli]

You can have more than one instance deployed, the caveat only one will be really active, others will have some level of idle.

But if the “master” fails, others will be able to pickup - they’ll elect a new master and they should continue to work l.

-

Alex

To view this discussion visit https://groups.google.com/d/msgid/kogito-development/79eb1dec-7832-4a15-a49e-93944d817466n%40googlegroups.com.

[Antonio Delli Carpini]

Hi Alex

the following software is present in the compact architecture (https://github.com/porcelli/jbpm-compact-architecture-example):

- PostgreSQL
- Management Console
- Task Console
- Keycloak
- PgAdmin
- Kogito Example Service

Normally for the election of a master there is a need for a coordinator (example Zookeeper), but in this case it does not exist.

I also saw at the aletyx project (https://docs.aletyx.ai/deployment/kubernetes/)
but at the moment the "Scaling and High Availability" section is empty.  (probably could be the key to this mystery)

Do you happen to have any suggestions?

Regards,

[Alex Porcelli]

The current master election is custom code within Job Service implementation relying on vert.x infrastructure.

Aletyx offers a commercial distribution, actually wihh the a different Job Service implementation, but given its commercial nature better discuss outside this mailing list that is focused on community. Feel free to reach out to al...@aletyx.ai

-

Alex

To view this discussion visit https://groups.google.com/d/msgid/kogito-development/e87fb8bc-eaca-4ad7-8f24-98a6cdc6c372n%40googlegroups.com.

[Antonio Delli Carpini]

Hi Alex,

while browsing online I came across the Aletyx documentation :) .

Thank you for the clarification.

 
I'm going back to Kogito open source.

So to recap:

there should be no problems using the compact architecture (https://github.com/porcelli/jbpm-compact-architecture-example) even with 2 or more runtime module containers. Even if there are N replicas, in reality only 1 will work. But in the event of disaster recovery, with the implementation of the job service which is based on the vert.x infrastructure, a new master will be automatically elected and therefore the system will work again.

Right?

Regards

[Alex Porcelli]

Sounds about correct

-

Alex

To view this discussion visit https://groups.google.com/d/msgid/kogito-development/5a9625e6-d9df-4e54-b693-f1777ab5f656n%40googlegroups.com.