Kubernetes Storage Developments That Enable Remote Work and Cloud-Native Applications

Original reporting source: forbes.com (May 2020). This article summarizes and expands on announcements made during that period.

Why Kubernetes Storage Matters for Remote Work

The rapid shift to remote work accelerated reliance on cloud-based applications—video conferencing, collaborative document editing, project management tools, and more. Most of these applications run in software containers hosted in large data centers. Kubernetes, the open-source container orchestration platform originally developed by Google and now stewarded by the Cloud Native Computing Foundation (CNCF), has become the de-facto standard for managing those containers at scale.

One persistent challenge with Kubernetes has been storage: containers are designed to be ephemeral, meaning they can spin up and shut down at any time. That works well for stateless workloads, but many real-world applications—databases, file services, analytics platforms—are stateful: they need to persist data across container restarts. Several vendors responded to this gap with notable announcements in early-to-mid 2020.

Background: Containers, Docker, and Kubernetes

A container is a lightweight, isolated execution environment that packages an application and its dependencies into a single unit. Unlike a virtual machine, a container shares the host operating system kernel, making it faster to start and more resource-efficient. Docker popularized the container format, and Kubernetes builds on top of it—and other container runtimes—to schedule, scale, and manage containers across clusters of machines.

• Containers provide isolation and portability across environments.
• Docker is a widely used container runtime and image format.
• Kubernetes orchestrates containers across multiple hosts, handling scheduling, health checks, scaling, and networking.
• Stateful applications require persistent storage so data survives container restarts or rescheduling.

Red Hat OpenShift: Unifying Virtual Machines, Containers, and Serverless

At its 2020 Red Hat Summit, IBM’s Red Hat announced several enhancements to OpenShift, its enterprise Kubernetes platform. Two announcements stood out for teams managing production Kubernetes infrastructure:

• OpenShift Virtualization: Allows organizations to deploy and manage virtual machines alongside containers and serverless workloads within a single platform. This is significant for enterprises that cannot immediately migrate all legacy applications to containers—they can run VMs and containers side by side under one management layer.
• Red Hat Advanced Cluster Management for Kubernetes: Provides a unified control plane for monitoring and deploying OpenShift clusters at scale across multiple environments, reducing operational complexity for teams managing distributed Kubernetes infrastructure.

Both features target a common enterprise pain point: hybrid environments where cloud-native workloads and traditional applications must coexist without requiring separate toolchains.

NetApp Project Astra: Enterprise Storage Services for Stateful Kubernetes Apps

NetApp introduced Project Astra as a software-defined storage and data services platform built specifically for Kubernetes, developed in collaboration with the Kubernetes community. Its stated goals include:

• Delivering enterprise-class data management (backup, recovery, snapshots) for Kubernetes-managed applications.
• Enabling application and data portability for stateful workloads—meaning applications and their data can be moved between clusters or cloud environments without data loss.
• Making it practical to run long-running, data-intensive workloads—such as databases—inside containers, which previously required complex workarounds to persist state reliably.

The core problem Project Astra addresses is that a standard Kubernetes pod, when deleted or rescheduled, loses any data written to its local filesystem. Persistent Volumes (PVs) and the Container Storage Interface (CSI) provide the framework; Project Astra aims to add higher-level data services on top of that framework.

Kioxia KumoScale: NVMe-oF Storage with Kubernetes CSI Support

Kioxia (formerly Toshiba Memory) enhanced its KumoScale storage software, which leverages NVMe over Fabrics (NVMe-oF) to deliver high-performance storage over a network. The relevant additions for Kubernetes operators include:

• CSI-compliant snapshots and clones: KumoScale now supports the Kubernetes Container Storage Interface standard for creating point-in-time snapshots and clones of volumes. This is important for backup workflows and for creating development or test copies of production data without disrupting running workloads.
• Reduced backup overhead: Kioxia states the snapshot and clone features reduce time pressure and storage consumption associated with backing up large stateful applications such as databases.
• Non-disruptive data cloning: Operators can clone production data for development, testing, or QA purposes without taking the production system offline.

NVMe-oF provides lower latency than traditional iSCSI or Fibre Channel storage networks, making it suitable for latency-sensitive stateful applications running in containers.

StorageOS v2.0: Software-Defined Persistent Storage for Kubernetes

UK-based startup StorageOS released StorageOS v2.0 into general availability as a cloud-native, software-defined persistent storage solution purpose-built for Kubernetes. Key features highlighted in the release include:

• Hyper-converged and multi-cluster support: Designed for organizations running Kubernetes across multiple clusters, whether on-premises, in the cloud, or in hybrid configurations.
• Encryption in transit: All traffic between storage nodes is encrypted and authenticated, addressing a security concern that becomes more pronounced when storage traffic crosses network boundaries in distributed environments.
• Cloud-native workflows: The storage layer is designed to integrate with standard Kubernetes primitives such as Persistent Volume Claims (PVCs) and the CSI, rather than requiring custom application changes.

Why These Developments Matter: Stateful Workloads at the Core of Remote Work Infrastructure

The applications that power remote work—file collaboration platforms, video services, databases backing SaaS tools—are overwhelmingly stateful. They write data that must persist and be recoverable. The Kubernetes ecosystem’s historical weakness in this area has been a barrier to running production-grade stateful applications in containers at scale.

The announcements from Red Hat, NetApp, Kioxia, and StorageOS represent a maturing of the Kubernetes storage ecosystem. Each addresses a different layer of the problem:

• Platform management (Red Hat Advanced Cluster Management) simplifies operating many Kubernetes clusters.
• Data services (NetApp Astra) provide backup, recovery, and portability for stateful applications.
• High-performance storage (Kioxia KumoScale + CSI) brings NVMe-oF speed to Kubernetes-native workflows.
• Software-defined persistence (StorageOS v2.0) delivers encrypted, portable storage for multi-cluster and hybrid environments.

Together, these developments lower the operational barriers to running complex, data-intensive applications in Kubernetes—supporting the cloud infrastructure that underpins modern distributed and remote workforces.

Key Concepts Reference