The generated protobuf files play a very important role in the Kogito platform. Each domain data from a custom Kogito service is generated as a protobuf when using the persistence add-on. These files need to be read by the Data Index in runtime.

On cloud environments, we leverage from the Kogito Operator to mount these files to a deployed Data Index instance. To do that, the operator needs to read the files from a given deployed custom Kogito service and mount to a deployed Data Index.

Today we read these files from the container's file system using shell script during the container initialization. This smells a anti-pattern since we are using a script inside the image to do that for us: expose the generated protobuf files.

Would be more interesting if the deployed service could serve these files through HTTP, so the operator would fetch them and update the Data Index volume.

This JIRA is to add these files to the META-INF/resources directory of the custom Kogito service, so the files would be served.