The latest version of Kubernetes goes a long way toward addressing fundamental networking and storage requirements that should serve to accelerate adoption of the container orchestration platform in production IT environments.
Version 1.11 of Kubernetes, scheduled to be available later this week, makes both IPVS-based in-cluster load balancing and CoreDNS as a cluster DNS add-on generally available.
Kubernetes 1.11 also adds alpha support for raw block volumes to the Container Storage Interface (CSI), along with an ability to integrate CSI with a new kubelet plugin registration mechanism and a simplified means for passing secrets to CSI plugins, noted the Cloud Native Computing Foundation (CNCF), which oversees the development of Kubernetes.
Other new storage capabilities include alpha support for the ability to dynamically resize persistent volumes without having to terminate pods and unmount a volume first and a facility that enables in-tree volume plugins to both specify the maximum number of volumes that can be attached to a node and limit them depending on the type of node.
The CNCF also announced the StorageObjectInUseProtection feature, which prevents the removal of persistent volumes that are bound to a persistent volume claim and persistent volume claims that are being used by a pod, is now stable.
Finally, there’s a beta version of a facility that administrators can employ to configure kubelets in a live cluster via the API server, and CustomResourceDefinitions are no longer restricted to defining a single version of the custom resource. That capability, also in beta feature, will be expanded in the future to support some automatic conversions. CustomResourceDefinitions now also support “status” and “scale” subresources, which integrate with monitoring and high-availability frameworks.
Stephen Augustus, specialist solution architect for the Red Hat OpenShift Tiger Team who also serves as the product management chair for the Kubernetes Special Interest Group (SIG) within the CNCF, says the advances in the realm of storage are especially significant as enterprise IT organizations move to build more robust applications that need to access persistent storage. As it becomes easier to manage storage within and without a Kubernetes cluster, the more feasible it becomes for traditional enterprise IT organizations to embrace Kubernetes, he says.
Kubernetes continues to evolve from being a powerful cluster built by engineers for engineers to a platform that is becoming more accessible to the average IT administrator. It may take a while for that transition to occur, but with each update Kubernetes is adding more both storage and networking capabilities and services that will make it more attractive to deploy Kubernetes in a production environment.
In the meantime, many more organizations continue to make extensive use of Kubernetes running on a variety of platforms to facilitate application development. Most of those instances are managed by a third party in the form of, for example, a cloud service provider or a third-party provider of a managed service. The real test, however, will come when IT organizations try to deploy and maintain Kubernetes in a production environment on their own.