I am running openFn via docker on ubuntu Ubuntu 24.04.3 LTS
Hi @mohau ! Thanks for reaching out! There was a bug (now patched in v2.15.12) where Kafka would show up as an option in this picklist even if you hadn’t configured a Kafka cluster and enabled the required settings in your .env file. (See the deployment section for details.)
Picking this option under these circumstances would, as you might expect, cause errors when saving your workflow! ![]()
If you update the version, Kafka should be removed from your picklist and you shouldn’t have any more troubles. (And if you do want to configure Kafka, it should work for basic use cases but we should note that this feature hasn’t been supported or used actively—as far as we know—in quite some time!)
If you’ve got a Kafka use case, please don’t hesitate to share it and we might be able to get it onto the roadmap soon ![]()
Hope this helps. And thanks for posting!!!
Thank you so much.
I actually have a use case that requires kafka.
I am trying to stream events from OpenLMIS through openFn (where we can then perform some transformations) then send them to Superset so that it can later be analysed. The openLMIS Reporting stack already has kafka listening for these events. I just want to injest them into OpenFn and then do some manipulation before sending them on their merry way.
