July 24, 2017

At the CoreOS Fest 2017 conference today, CoreOS outlined a hybrid cloud computing strategy based on Kubernetes clusters deployed on public clouds and on-premises.

CoreOS CEO Alex Polvi says one of the reasons hybrid cloud computing has not been achieved is that there’s no real platform commonality between on-premises and in the public cloud. Most internal IT organizations, for example, rely on VMware virtual machines while public cloud providers tend to favor open-source virtual machine. The result has been a bifurcated approach to enterprise IT with multiple clouds all managed in isolation.

By making an available a Tectonic instance of Kubernetes managed by CoreOS as a cloud service, IT organizations now can easily move workloads between on-premises environments running Kubernetes and the cloud, Polvi says. That capability should enable IT organizations for the first time to deploy and move workloads across multiple regions in a cloud, for example.

CoreOS also announced that Container Linux Operator, a set of automation tools developed by CoreOS, will include Tectonic version 1.6.4 later this year. In addition, CoreOS unfurled etcd as a service, which makes it possible to automatically deploy key/value data stores on a cloud. The etcd Operator handles all etcd scaling, failure and version updates.

A new survey of 200 IT decision makers from 451 Research commissioned by CoreOS finds that 57 percent of organizations prefer a combination of both hosted containers as a service (CaaS) and on-premises.

The survey also finds that 71 percent of respondents indicated they are using Kubernetes with more than half of organizations (52 percent) indicating they are running container management and orchestration software in production today. Top benefits cited in the survey are the freeing of resources to focus on other priorities, hybrid cloud/cross-cloud support, the ability to bring your own compute and avoiding vendor lock-in.

The rise of Kubernetes and other container orchestration platforms is having a profound effect on cloud computing strategies. Once an organization standardizes on Kubernetes, for example, it becomes easier to play one cloud service provider off another. If an IT organization doesn’t appreciate the terms or price of a service, moving a workload from one instance of Kubernetes to another is a lot less trivial than moving workloads between different types of virtual machines.

Of course, most containers in production environments at least still run on top of virtual machines. But, as Kubernetes becomes more robust with each passing day, many IT organizations are opting to deploy Kubernetes without needing any assistance from cloud frameworks such as OpenStack or VMware. Standing up a Kubernetes cluster is still not as easy as it might be. But as the number of managed services based on Kubernetes continues to proliferate, much of the overhead associated with deploying and managing Kubernetes becomes someone else’s problem.

Obviously, it’s not likely most organizations will replace their legacy infrastructure with platforms such as Kubernetes overnight. But for many of them, a cloud-first strategy may very well evolve into a Kubernetes-first strategy that promises to provide a lot more in the way of hybrid cloud computing flexibility.

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.