Introduce an Embedded Data Base to Serverless Workflow applications

For a quick and fast test and development iteration in a Kubernetes cluster, the Kogito Serverless Workflow applications should be able to be deployed using an embedded database as an alternative to a more complex server, such as Postgres.

This persistence alternative should be easy to configure and set up. Ideally, via an external configuration where external agents could set. For example: my.db.path=/home/kogito/db.

Additionally, keep in mind that this file system can potentially be shared among many other pods running the same workflow instance.

This should be an alternative for non-production environments only.

The outcome of this EPIC is:

• A study of viability for a potential embedded db
• A benchmark for multiple writes and reads from different instances of the same workflow application
• The implementation as a Kogito Quarkus Extension Add-on of the chosen database.
• A PoC of a workflow application with embedded persistence running in a Kubernetes cluster scaling to a few pods (and to zero) with Knative.