Problem:

The upstream effort is coming to a critical junction with the SBO spec 1.0 GA approaching. There are a few issues that are considered blockers for 1.0 to be GA. This epic aims to reserve some cycle for this work as the specs compliance will give us more credibility and potentially catapult more adoption. Another aspect of the upstream work we are doing with IBM and VMWare is the reference (core implementation) this implementation will help use create a true upstream community with extensibility needed to create our SBO 2.0. The needs to be a considerable investment upstream in terms of core implementation as well as conformance test that can run on a specified implementation.

Goal:

Release spec 1.0 GA and implement an extensible core implementation to GA as 1.0 as well. Like OpenShift does with Kubernetes the reference implementation will need to be something we can build on top, otherwise, it will become a costly duplicate with dual maintenance required on our SBO codebase.

Why is it important?

Build a true multi vendor community with higher probability of adoption. Build a partnership with other vendors that can help maintain the core implementation.

Use cases

1. Be able to reference a stable 1.0 spec
2. Be able to use Reference implementation 1.0 as the base for our SBO 2.0

Acceptance criteria

1. We have a 1.0 reference implementation
2. We have a feature complete reference implementation the can be extended with additional feature.
3. We have done Performance testing on Reference Implementation
4. Conformance test is done.

Dependencies (External/Internal)

Design Artifacts

Looks like we already know what needs to be done, we have one VMWare engineer willing to spend the time. We need to either match that or be clear about our strategy otherwise.

Exploration

Note