OCI Enables Persistent File Storage on Serverless OKE Virtual Nodes

Oracle Cloud Infrastructure (OCI) has introduced support for persistent storage on Oracle Cloud Infrastructure Kubernetes Engine (OKE) Virtual Nodes via OCI File Storage service (FSS). By integrating with the OCI Container Storage Interface (CSI) driver, OKE Virtual Nodes can now provision and attach persistent volume claims (PVCs) dynamically or statically using standard Kubernetes primitives. The capability includes full ReadWriteMany (RWX) access modes, enabling multiple serverless pods across fault domains to mount shared, POSIX-compliant file systems with directory-level granular controls. Virtual Nodes provide a serverless operating model for Kubernetes by delegating operating system management, host provisioning, scaling, and node patching entirely to Oracle. However, until this integration, Virtual Node workloads were confined to ephemeral local storage, preventing teams from deploying stateful applications without reverting to self-managed or managed node pools. Supporting FSS-backed storage extends serverless execution to complex workloads such as shared content management backends, collaborative data-processing pipelines, CI/CD runners with shared build artifacts, and AI model checkpoints. Platform operators gain complete decoupling of compute scaling from state persistence without maintaining underlying worker machines. This update mirrors a wider industry movement across hyperscale cloud providers—such as AWS Fargate with Amazon EFS and Google Cloud GKE Autopilot with Filestore—to bridge the historical chasm between serverless container runtimes and persistent storage. As enterprise containerization matures, organizations increasingly prioritize operational simplicity without sacrificing architectural flexibility. Providing native RWX file storage within serverless Kubernetes removes one of the final blockers for migrating legacy enterprise patterns and modern data pipelines into fully managed operational frameworks. For platform engineers and cloud architects, designing stateful deployments on OKE Virtual Nodes requires minimal changes to existing CI/CD workflows or Helm charts. Teams should define an FSS StorageClass and configure appropriate subpaths to isolate application directories on shared mount targets. However, teams must account for key architectural constraints: FSS capacity charges and mount target bandwidth limits operate independently of Kubernetes pod resource requests. Additionally, cluster operators should configure appropriate reclaim policies (such as Retain) to protect critical data during chart lifecycle upgrades and ensure subnet security lists permit NFS traffic between the Virtual Node subnets and FSS mount targets.