Camunda BPM and Database

I understand Camunda BPM works only with Database, but just wondering if there is any workaround : like if anyone tried to parse BPM XML, similar to DMN Java Main example. I’m trying to use limited functionality to just orchestrate the services and rely on existing Splunk logging to capture the historical information.

Even if there is no other way around, it works even if we can filter and ignore certain tables, keeping the ones that just stores the definitions. I’m currently trying to use just the engines for this year, while we evaluate and on board to web app and database etc early next year.

Few of the concerns that keeps hitting us is how much space is required in the database for Camunda to store history and all the activities, how frequent the data gets cleaned up or do we have to introduce archive scripts or some kind of house keeping scripts etc.

Any thoughts or advise helps.

Hi, camunda can work without DB, peoples from Goldman Sachs use in this case.

Also you have some options to control your history size:

  1. history time-to-live - you can configure it to 1 day min.
  2. use rest to clean up -https://docs.camunda.org/manual/7.9/reference/rest/history/process-instance/post-delete/