OpenStack, OpenShift and the Recovery Gap Between Layers

OpenStack is the infrastructure layer. OpenShift is the application platform layer, built on Kubernetes. The two are designed to work together, and neither is meant to replace the other.

Combined with everything gathered so far, the sources describe one consistent model with four layers:

Layer 1, infrastructure: OpenStack handles compute, storage and networking.

Layer 2, application platform: Kubernetes or OpenShift deploys, scales and orchestrates apps.

Layer 3, application and data: databases, services and their dependencies.

Layer 4, tooling around it: operators, backup tools and automation.

Each layer is separate, optimized independently and owned by a different team, while the system itself is a combination of all of them. Every document supports this: modern architecture is layered by design, and each layer only guarantees its own responsibility. As a result, no layer guarantees that the full system works together.

The headline is that modern cloud platforms are built in layers, but failures don't happen one layer at a time. Infrastructure runs on OpenStack, applications run on OpenShift or Kubernetes, and data lives in databases. Each layer works independently, yet the system fails as a whole, and that is the gap.

The strongest version is “Your platform works in layers. Your failure doesn’t.” A cleaner alternative is “Everything is designed in layers. Recovery is not.”

This version holds up because it is grounded in all the sources and consistent across architectures. It doesn't attack any tool, it points to a structural gap, and it is easy to understand. Together, the material covers layer separation, responsibility separation, tooling separation and the recovery gap.

The remaining work is to use it. Take this line, “Your system is layered. Your failure isn’t.”, and test it with real buyers to validate it.