Camunda Kafka Orchestration Error Handling

[mattijt]

We are using Camunda to orchestrate a order integration flow with microservcies listening to a kafka topic.

This is the Camunda pattern:

The Transform Order Task will place a message on the kafka topic to trigger an external service to process a specific order. Once the external service is complete with the transformation, it will place another message back on the kafka topic for the Order Transformed task to pick up and the flow will continue.

How would I create an incident on the Transform Order task if the external service throws an exception or otherwise fails? I would like this to be a process where we can retry the Transform Order task, to reinitiate the processing of the order. If the external service throws an exception, it puts a message on a DLQ topic which the Camunda service we built is monitoring.

[jonathan.lukas]

Hello @mattijt ,

to achieve thsi, the worker handling the job would have to wait with completion until the transformation was tried. It could then complete the task or fail the task.

Looking at the async pattern, this does not make a lot of sense here. How long will a transformation take? Would it make sense to model it sync (with a simple service task) to achieve the desired behaviour?

I hope this helps

Jonathan