Do we want to make a recommendation for greenfield projects (default architecture)

Seeing @javahippie’s post here I’d like to discuss if we should recommend a default architecture for new greenfield projects or not.

We have a couple of options:

  • recommend the external task pattern
  • recommend embedding the engine
  • recommend nothing in particular and show the two options side by side
  • recommend nothing

What do you think our policy in this regard should be?

My strong opinion is, that the flexibility of Operaton is one of its major strengths. Camunda recommended the standalone engine / external task pattern heavily in the recent years, but its main purpose was to prepare customers for a C8 migration.

As the different architectural styles have different strengths, I would be very hesitant to recommend any of them as the default, as it might raise the impression, that it performs worse in other settings, when it’s just a matter of

  • Do you need transactional services?
  • Do you need to integrate different languages?
  • How does your load and processing power look like?
1 Like

I agree that the flexibility is a core strength. When I first came in touch with C7, I liked the ease to integrate this into SpringBoot. But thinking more about that I thought that external tasks are powerful to orchestrate processes with a central engine and low-code. So both are totally valid.
So I would like to target also both: I would like to have show cases which allow easy scaffolding of applications and services on the one hand, and maybe on the other hand having a Operaton instance running on a cloud provider where users could deploy a (limited) number of processes and interact with external services via REST or messages.