IBM announces managed Kubernetes on bare-metal servers
IBM announced its Cloud Container Service, a fully-managed container services based in Kubernetes, is now able to run on bare metal nodes.
Containers are a lightweight form of virtualization, which enable developers to run an application and its dependencies in an isolated environment. They have become a popular way to run multiple, isolated applications on a single server.
Nevertheless, a drawback of running containers on bare metal is it usually requires configuration and constant management by development teams. Managed services makes the process easier with benefits like automatic updating, intelligent scaling and built-in security.
With IBM’s offering, the benefits of containers can be applied to data-intensive workloads that demand high-computing performance, including large machine learning workloads and sensitive datasets that need to remain isolated. According to Jason McGee, vice president and CTO for IBM’s cloud platform, running the IBM Cloud Container Service on bare metal nodes, “gives developers greater control over where their workloads reside and enables them to isolate workloads to specific servers,” he wrote in a company blog post. “It equips teams with all of the benefits of a fully managed container service combined with the performance and security of bare metal.”
Kubernetes is an open source platform recognized as the de facto standard for container orchestration. Major cloud providers have enhanced their offerings by embracing Kubernetes. Last November, for example, Amazon Web Services (AWS) added Kubernetes to its Elastic Container Services (EKS). The move followed Microsoft’s release of dedicated Azure Container Service (ACS) for Kubernetes the month before. According to McGee, however, IBM is the first major cloud provider to offer Kubernetes containers as a managed service directly on bare metal cloud infrastructure.
McGee added that IBM is working with the open source community on offering applications created by Kubernetes to directly access graphical processing units (GPUs). GPUs are computers chips that perform mathematical calculations required to render and display images. They serve as the “make-or-break component of many cloud-enabled technologies,” according to McGee, like machine learning, which are data and compute-intensive.
In IBM related news, the company announced it is working with web security company CloudFlare. As part of the partnership, IBM becomes an authorized reseller of Cloudflare’s suite of security and performance edge services. IBM customers will be able to leverage Cloudflare’s services usually by clicking a button in the dashboard.