At the moment i am testing and thinking about the usage of Camunda.
My long term goal is to have a DDD architecture with different micro services and having a workflow (Camunda) which uses them to execute a flow.
I read this article
And i was wondering if anyone tried using the following concept and maybe also using the following project.
Update :
For now i will be following the approach which is being detailed in the article.
I will add a seperate Camunda Engine and a seperate c# camundaClient project (with adjustments) as windows services for each bounded context.
@StephenOTT Yeah, I decided to follow the external task approach.
Because so far it seems to me that,
This way the BPMN diagram is “code free”,
I can control the tasks better ( failure, complete etc )
I can split the workers load into more servers if need to
But i guess i will see how well it works out as i continue to dig deeper into the implementation.
Another pattern we have followed is to build rest based micro services. Then use camunda to pass data to different endpoints and manipulate data using scripting in camunda (JavaScript/nashorn). But other script languages would be available as well: https://docs.camunda.org/manual/7.7/user-guide/process-engine/scripting/
Well i am hoping to be able to keep the BPMN as minimal as possible in terms of “developer work”.
An ideal situation will be to have “domain experts” work with the camunda modeler to create flows and the developers having later minimal work with integrating it with the Code behind ( C# Services etc ).