How does a deployment aware job executor pick jobs?


I’ve read this thread but still don’t seem to be able to answer the following question.

Assume the following scenario:

  1. We create a docker container with a camunda app; the engine is configured as “deployment aware”. The app contains a process model with the key “TheProcess” and version tag “v1.0”.

  2. We start some instances of the process. The instances reach a waiting state, e.g. an intermediate catch event. It’s only important that there are some async job entries that “belong” to this container.

  3. We remove the docker container and create a new one. The new one contains the camunda app, but now it’s a slightly modified version of the process. The process key is the same, the version tag is changed to “v1.1”. The process model with “v1.0” is not in the container anymore (but it’s still in the process DB).

The question is: Will the new container (the job executor in it) pick up the async jobs left behind the first (no more existent) container? I.e. will the process instances started by the old container be further executed by the new container? We can assume that all the bean references in the process are still resolvable.

Thank you for the clarification!