Mirantis this week at the KubeCon + CloudNativeCon North America 2025 event, added an ability to manage virtual machines to an open source kordent control plane it originally developed to manage Kubernetes clusters.

Company CTO Shaun O’Meara said the Mirantis k0rdent Virtualization extension for virtual machines that has been added to the Mirantis k0rdent Enterprise edition of the platform will enable IT teams to unify the management of legacy and cloud-native infrastructure.

Based on the Cluster application programming interface (API) in Kubernetes, the kordent control plane has now been integrated with virtual machines to enable providing an Automated Distributed Resource Balancer (DRB) for virtual machines that enables IT teams to identify servers with sufficient CPU and RAM resources needed to rebalance workloads.

Additionally, Mirantis has extended its support for graphical processor units (GPUs) to enable kordent to preserve non-uniform memory access (NUMA) and PCIe topologies of virtual machines to enable IT teams to optimize application performance.

A hugepages and CPU-pinning capability is also being enabled to improve performance via efficient memory management and the allocation of dedicated processor resources.

Finally, Mirantis has also added support for the latest edition of kubevirt, an open source tool for encapsulating virtual machines in containers, and the Ceph storage management platform.

Control planes were initially developed by cloud service providers to manage highly distributed computing resources. Mirantis developed kordent to provide internal IT teams with that same capability. Many of those teams are now increasingly managing highly distributed workloads, especially as IT teams move to comply with sovereign cloud requirements that result in more applications being deployed in on-premises IT environments, noted O’Meara.

Additionally, control planes present an opportunity to reduce the total cost of IT by eliminating the need to have separate management platforms for virtual machines and Kubernetes clusters.

It’s not clear how many IT teams are moving to unify the management of IT infrastructure, but as the number of workloads deployed continues to increase, there is a growing need. Many organizations simply can not afford to hire the additional IT staff that, in the absence of a unified control plane, would be needed to manage those workloads.

The number of workloads running on virtual machines, Kubernetes clusters and bare-metal servers, of course, varies widely from one organization to the next. Many IT teams are currently in the process of modernizing some of those workloads as part of an effort to refactor them to run on Kubernetes clusters. At the same time, the number of bare-metal servers configured with GPUs to run artificial intelligence (AI) workloads is also starting to increase.

As the types of workloads being deployed become more diverse, managing IT is only going to become that much more challenging. Hopefully, with the rise of AI, it will become simpler to manage those workloads, but even with those advances, there is still a need for a control plane that makes it possible to streamline the management of IT in a way that ultimately serves to make IT teams more agile.