RedHat OpenShift is a container platform built upon and extends the Kubernetes functionality. In addition to Kubenetes' ability of orchestrating containerized applications, OpenShift provides support for the complete CI/CD life cycle of applications, called Source-To-Image.

The API Gateway OpenShift support provides the following, in the same way as the Kubernetes support does:

OpenShift extends Kubernetes and introduces new objects. For example, Kubernetes deployment is called DeploymentConfig and has the version id apps.openshift.io/v1 . In order to make services accessible from outside the cluster, OpenShift provides Route objects. The images required to start containers are not necessarily referenced directly inside the container specification, rather they can be managed by ImageStream objects.

OpenShift has a specific way for running ElasticSearch containers. ElasticSearch needs an increased virtual memory mmap count: vm.max_map_count >= 262144. In a plain Kubernetes environment you can solve this by adding an initContainer that has to run in the privileged mode. OpenShift offers a much simpler solution. If a pod carries a specific label then OpenShift applies the necessary system changes behind the scenes when starting the pod's containers.

For details on how these OpenShift specific topics are reflected in YAML configuration files for API Gateway, see OpenShift Sample Files.

When starting a new container, by default, OpenShift ignores the built-in user of the Docker image and injects a new user. This user is a member of the root group, and hence the files, scripts, and programs inside the container have to be readable, writable, and executable by the root group. To understand how to work with this OpenShift behavior, see the following sections: