On 30 January Proteon attended OpenShift Commons in London. Just recently OpenShift 4.0 was announced and we wanted to get updated by the core team what was going on. Many thanks to Diane Mueller for organizing OpenShift Commons and her effort to get the key people of the OpenShift team to Europe.
The day started off with an introduction of the new features of OpenShift 4. The main new features of OpenShift 4 are:
- Autoscaling clusters: OpenShift clusters can scale as easy as applications can scale now. Until now autoscaling Kubernetes clusters was only possible with Cloud vendor specific constructs. While the core of this feature is vendor agnostic, only AWS will be supported in the first release.
- Better manageability thanks to CoreOS, which means better security at lower operational cost.
- Better security, especially for stateful (legacy) applications, which means more applications can run on OpenShift and enjoy auto recovery and autoscaling features.
- Extensibility through the Operator Framework. OpenShift 4 allows technology companies to offer OpenShift extensions, both free and paid through a Marketplace. This is important for technology providers and OpenShift users alike because it allows a marketplace without vendor-lock-in by a particular Cloud provider.
- The web console has been replaced with a new version based on CoreOS’ Tectonic offering more complete and structured access to all OpenShift features.
After this, Mike Barrett, OpenShift Product Manager, shared the OpenShift 4 roadmap. OpenShift 4.0 will be launched in April 2019, just in time for the leading Kubernetes event: Kubecon Barcelona. However, we are a bit surprised that OpenShift 4.0 will only support AWS installations because the strength of OpenShift should be platform independent. So we consider OpenShift 4.0 as the pre-release version. OpenShift 4.1, due in May 2019 will be infrastructure independent. It will need configuration and customization on the provisioning Operators, but that is exactly what Service Providers such as Proteon would handle. Also, May 2019 is the release time for complete, updated documentation. Another important aspect of this is how to update existing clusters to 4. OpenShift will not offer an in-place upgrade path, because the fundamental changes (Red Hat CoreOS will replace RHEL and to
So, what does this mean for OpenShift users? If you have a cluster prior to 3.9, you should upgrade to 3.9 at least. If you have a 3.9 or 3.11 cluster you can stick to that until you need the OpenShift 4 features. If you want to get started with OpenShift: start with 3.11 and upgrade later.
Proteon has already started internal tests with OpenShift 4.0. If you want a preview of OpenShift 4.0, please do not hesitate to get in touch!