Latest Kubernetes 1.27 Release Provides More Control

The latest version 1.27 of Kubernetes adds a range of capabilities that promise to give IT teams more granular control over pods within individual clusters.

Xander Grzywinski, release lead for this version and a senior product manager for Microsoft, says the capabilities that stand out most are an instance of in-place vertical pod autoscaling that is available in alpha and, in beta, the ability to schedule when a pod can be considered for scheduling and support for mutable pod scaling that might make it possible to one day employ lighter weight controllers.

In addition, a read-write once access mode that restricts access to a storage volume to a single pod and an alpha node log accessed via the Kubernetes application programming interface (API) to make it easier to debug services while running is now in beta. The volume manager, also available in beta, has been refactored to allow the kubelet to populate additional information about how existing volumes are mounted during startup.

Those features will add additional capabilities for deploying stateful applications that are now being deployed more frequently on Kubernetes clusters, notes Grzywinski.

Other capabilities that are now generally available include mutable scheduling directives that makes it possible for a custom queue controller to decide when a job should start and a Respect PodTopologySpread after rolling upgrades capability.

In general, there is a growing recognition that Kubernetes is more than just a container orchestrator; it provides IT teams with a full-fledged approach to declaratively managing resources, says Grzywinski.

At the same time, it’s becoming simpler to provision Kubernetes. In fact, many IT administrators can now provision it without having a lot of DevOps skills, he notes.

The challenge IT organizations encounter is that policy and security management often take a back seat to raw resource management, which Grzywinski notes can lead to a lot of issues down the line.

One of those issues is the number of Kubernetes versions organizations are running. Many organizations are hesitant to upgrade Kubernetes clusters for fear of breaking applications. There is a fierce debate within the community about creating a Long Term Release edition of Kubernetes that, however, remains unresolved.

Kubernetes, of course, has never been more popular among enterprise IT organizations that are employing deploying multiple cloud-native applications in production environments. The issue that is arising is that not all those applications are fit for Kubernetes purpose. Many of them, for example, might be better suited to run on a serverless computing framework.

It may be a while before cloud-native applications supplant monolithic applications in the enterprise but they are becoming more prevalent every day. With that change comes a range of management challenges that become easier to address with each subsequent update to Kubernetes. The challenge, of course, is that too many versions of Kubernetes become too much of a good thing—making an already complex platform even more difficult to manage at scale.

Mike Vizard

Mike Vizard is a seasoned IT journalist with over 25 years of experience. He also contributed to IT Business Edge, Channel Insider, Baseline and a variety of other IT titles. Previously, Vizard was the editorial director for Ziff-Davis Enterprise as well as Editor-in-Chief for CRN and InfoWorld.

Mike Vizard has 1615 posts and counting. See all posts by Mike Vizard