Introducing Nutanix Flow CNI

A unified network for VMs and containers

Enterprises are entering a new era where applications run across multiple forms of infrastructure at the same time. A large portion of critical workloads continue to run on virtual machines for stability, operating system control, and commercial software support. At the same time, teams accelerate delivery by pushing frontends, APIs, and microservices into Kubernetes® container management. Physical servers remain essential where performance and hardware level control matter.

This mix is now the default model. A single application often spans containers, VMs, and physical systems. A frontend API may run as pods. A payment engine may stay on VMs due to compliance. A latency sensitive database may run directly on hardware. For the application to function end to end, all these components must communicate cleanly and reliably. This requires a network that can act as a federated VPC across these different execution environments, so workloads can operate together even when they run on separate platforms.

The challenge is that pod networks and VM networks have always lived in separate worlds. They follow different routing models, require additional gateways or load balancers, and force teams to operate two networking stacks. This can add delay, increase operational noise, and make application design more complex than it should be. Enter the Nutanix Flow CNI solution - a Container Network Interface for today’s modern architecture.

Nutanix Flow CNI changes this.

It brings a single networking model for both pods and VMs. Pods become first class citizens in the Nutanix VPC routing domain. VMs and pods share one structure, one path, and one set of behaviors. This unified behavior is what enables true workload federation across the platform. Communication becomes direct, predictable, and free from the extra components that slow down mixed applications today.

This is not a small improvement. It is a shift in how Kubernetes and VM workloads can be connected inside Nutanix environments.

Why a unified model is needed

Most Container Network Interface (CNI) solutions were built only for Kubernetes clusters. They provide pod to pod connectivity, but they stop at the cluster boundary. They do not extend awareness to the broader environment that customers rely on every day. This leads to familiar challenges:

Split networks

Pods and VMs live in different domains and require gateways to talk to each other.

Extra hops

Frontends in Kubernetes must cross several layers before they can reach VM based services.

Operational complexity

Each domain has its own routing logic, troubleshooting tools, and operational paths.

Application design impact

Architects must take the network split into account even when building simple services.

These limitations appear in every industry. Retail platforms pair containerized web layers with VM based payment engines. Banks modernize APIs on Kubernetes while keeping regulated systems on VMs. Telecom and edge deployments combine Virtual Network Functions (VNF)s on VMs with modern microservices running as pods. In all cases, the network stack may end up as the slowest and least flexible part of the architecture.

Flow CNI removes that divide.

What Flow CNI brings

Flow CNI removes the long standing separation between Kubernetes and VM networks. With Flow CNI, teams no longer need to think about whether an application tier should run as a VM or as a container. Both forms become native citizens of the same VPC and follow the same networking domain. This removes one of the oldest constraints in hybrid application design. A service can stay on a VM, move to Kubernetes, or run across both without any change in the networking layer. This is a natural step toward a federated VPC model where applications span multiple environments without introducing new network boundaries. This is the practical meaning of running anything anywhere inside a Nutanix environment. Flow CNI gives enterprises the freedom to modernize at their own pace without redesigning networks or splitting architectures across technology boundaries. It brings one connected fabric that supports every form factor, everywhere it runs.

One VPC for all workload types

Pods receive native VPC IP addresses, just like VMs. There is no longer a boundary between Kubernetes and the rest of the infrastructure.

Direct communication

Pods connect to VM based services without external gateways or load balancers. The communication path becomes short, efficient, and easy to understand.

Consistent routing

All workloads follow one routing model. Operators troubleshoot with a single view.

Fewer moving parts

Flow CNI removes layers traditionally required to connect Kubernetes to VM environments. This design reduces failure points and helps to simplify deployment and speed the path to production.

Support for real hybrid applications

Companies often modernize one layer at a time. They move frontends to containers and keep core logic on VMs. Flow CNI supports these architectures without forcing full redesigns.

A practical example

Consider a simple three tier application:

• The frontend runs as pods on Kubernetes
• The business logic runs on virtual machines
• The database runs on a physical server

With Flow CNI, all tiers join the same VPC. Communication stays inside a unified routing domain. There is no load balancer or gateway required between pods, VMs, or physical hosts. Even in this simple example, the workloads behave as part of a federated environment, despite running on different platforms.

This reflects how you evolve your applications today: gradually, pragmatically, and without rewriting everything at once.

What this means for enterprise environments

Flow CNI introduces a clean and unified networking experience for mixed workloads. It removes the long standing split between Kubernetes and VM networks and eliminates the operational complexity that comes with it. It gives companies a single structure that supports incremental modernization at their own pace, while delivering performance and simplifying operations.

For the first time, Nutanix environments gain one consistent networking model for containers, VMs, and physical workloads.

These capabilities form the basis of a broader workload federation model that can bring consistency to any environment where applications are distributed across different tiers. This is the foundation for the next phase of application connectivity across Nutanix platforms.

See what Flow CNI can do for you

This is the first release of Flow CNI, and it marks a significant step forward in how you connect Kubernetes and virtual machine environments on Nutanix. If you want to understand how this new model can simplify your architecture and operations, here are the best ways to dive deeper:

• Visit the public documentation for full details and examples
• Contact your Nutanix representative for a technical walkthrough or a live demonstration
• Request a dedicated session with our product team to review your use cases and discuss how Flow CNI can support your environment

©2025 Nutanix, Inc. All rights reserved. Nutanix, the Nutanix logo and all Nutanix product and service names mentioned are registered trademarks or trademarks of Nutanix, Inc. in the United States and other countries. Kubernetes is a registered trademark of The Linux Foundation in the United States and other countries All other brand names mentioned are for identification purposes only and may be the trademarks of their respective holder(s).