The world of Kubernetes storage is experiencing a transformation that goes far beyond what most people expect, with virtual machines, AI inference, and mission-critical databases all converging on cloud-native platforms.
In this conversation from KubeCon and CloudNativeCon Europe, Greg Muscarella, general manager at Portworx (EverPure formerly known as Pure Storage), shares insights that might surprise even experienced Kubernetes practitioners. The discussion explores why hundreds of thousands of virtual machines are now running on Kubernetes, what technical innovations enable this, and how storage requirements differ from those of traditional containerized workloads.
Watch this video to discover the hidden complexities of storage in modern Kubernetes environments and learn why major enterprises are making strategic platform decisions that seemed impossible just a few years ago.
Storage for Kubernetes
Muscarella explains the technical reasons why Kubernetes storage has become dramatically more sophisticated. You’ll understand why the original assumption that Kubernetes was only for stateless applications is no longer valid, and what that means for infrastructure architecture decisions.
The conversation explores a trend that’s reshaping enterprise infrastructure: massive-scale virtual machine deployments on Kubernetes. Muscarella explains the specific storage challenges this creates, problems that don’t exist with traditional containerized applications, and how Portworx solves them. The discussion includes real-world examples of manufacturing organizations running their entire operations on this architecture.
Key topics covered
- Why virtual machines on Kubernetes are becoming common at enterprise scale
- The technical difference between storage for containers versus VMs
- How live migration works differently in Kubernetes environments
- What read-write-many block storage enables for virtualized workloads
- The role of Kubevirt in enterprise Kubernetes deployments
- How disaster recovery and backup work across distributed locations
- Why the Broadcom/VMware situation accelerated Kubernetes VM adoption
- The importance of control and platform independence for enterprises
- How AI inference workloads are impacting storage requirements
- What open source contribution means beyond just writing code
- The future trajectory of the Kubernetes ecosystem
From VM to Kubernetes
The conversation reveals that the Kubernetes ecosystem has reached an inflection point. What started as a platform for simple containerized applications now handles workloads that were traditionally the domain of expensive, proprietary virtualization platforms. Understanding how storage enables this transition is critical for anyone making infrastructure decisions.
Muscarella shares candid insights about the balance between proprietary software and open source community contributions, explaining how Portworx navigates this landscape while remaining fully committed to Kubernetes ecosystem growth. The discussion also covers how different workload types like: databases, Kafka pipelines, and financial services applications require different storage characteristics and how modern platforms accommodate these needs.
Whether you’re an infrastructure architect evaluating platform options, a developer working with stateful applications, or simply curious about where cloud-native technology is heading, this conversation provides valuable perspective on the technical and business factors shaping Kubernetes adoption at enterprise scale.