Release

Red Hat releases new minor versions of OpenShift Container Platform approximately once every 4 months. OpenShift versions use a Semantic Versioning style of denotation to designate specific versions. For example, version "4.8.1" of OpenShift would designate that the major version is "4", the minor version is "8", and the patch level is "1". OpenShift is currently releasing along the major version "4" series. When the next version of OpenShift is released, it will be "4.n+1", where "n" is the current minor version. For example, at the time of writing the current version of OpenShift is "4.9", the next release will then be "4.10".

Z-Stream Releases

While new minor versions of OpenShift are released every 4 months, "Z-Stream" releases are created much more frequently to address security and critical bug fixes. A z-stream release is designated by an increase in the patch level version of the release. For example, version "4.8.1" of OpenShift would designate that the current patch level is "1", if a new z-stream release is needed it will be version "4.8.2".

In general, backports to older versions of OpenShift will be evaluated on a case-by-case basis. Security and critical bug fixes will be considered the highest priority for backporting. Features are not backported without a justified exception from Red Hat.

Integrating with the Release Process

For any minor version of OpenShift the release process consists of several important phases which are traversed in this manner:

Feature Development -> Feature Freeze -> Code Freeze -> General Availability

Feature Development

The feature development phase is a time when any change can be proposed to the code repositories. There are no automated limitations on the pull requests that can be proposed, and any work related to new features can be considered for inclusion. Ideally, the bulk of new work will be created during this phase of development.

Feature Freeze

Feature freeze is an important concept to understand when integrating with the Red Hat release process. During the development of a new version of OpenShift there will come a time when new feature work will be disallowed into the code base. This process is referred to as "feature freeze" and it is used to create a hard deadline for adding new features to a release. As new infrastructure platforms are considered features, new components and changes should be in place before the feature freeze.

Approximately 12 weeks before the general availability release for an OpenShift version, the engineering team will observe the feature freeze. After the feature freeze date, new pull requests will require a related bugzilla issue in order to be considered for merging. This is not a mechanism for bringing features in though, and should only be used to fix bugs within the existing code.

What this means is that new infrastructure providers should work to ensure their provider specific components have been created and included in continuous integration before the feature freeze. There will still be time to fix bugs, but the major pieces should be in place (e.g. repositories created, continuous integration configured, image builds automated).

Code Freeze

Approximately 6 weeks after feature freeze the OpenShift engineering team will observe a "code freeze". The code freeze indicates that no more pull requests will be taken from the source repositories, and that the release teams will now take over the process of performing the final testing and analysis, and packaging of release artifacts.

During the time between the feature freeze and the code freeze, new pull requests will be accepted provided they have a related bugzilla issue. After code freeze these changes will not be considered or accepted unless they are release blocking issues.

General Availability

Once the source base that was compiled during the code freeze has been fully tested and vetted by the Red Hat release teams, the new OpenShift Container Platform version will be released for general availability. This process is usually predicted to occur 6 weeks after the code freeze, but given the possibility of release blocking bugs this is only an approximation.

If you have followed the instructions in this guide then your new components will be ready for inclusion with the next OpenShift minor release. The final piece necessary for being included in an OpenShift release from Red Hat are full end-to-end integration of your infrastructure platform into the continuous integration process (this is covered in greater detail in Continuous Integration and Testing).

When the end-to-end integration tests of the new infrastructure platform are passing regularly, the new platform can be included in an OpenShift release. The new components will have already been included in nightly and continuous integration payloads, and they will now be promoted to the release payload.

Release Life Cycle

After release, Red Hat provides a support and maintenance life cycle for the version of OpenShift Container Platform depending on the release type and version number. For more information about this process and the dates associated with extended support, please see the Red Hat OpenShift Container Platform Life Cycle Policy document.

For some historical context around the release schedule and how Red Hat has arrived at this process, please see the following articles: