→ Back to Home
Kubernetes

Kubernetes v1.36 Now Available on OCI Kubernetes Engine (OKE)

Oracle Cloud Infrastructure (OCI) has officially rolled out support for Kubernetes v1.36 on its Kubernetes Engine (OKE), offering customers immediate access to the latest upstream Kubernetes features. This update, announced today, May 28, 2026, is designed to deliver enhanced security, more efficient resource management, and a simplified operational experience for containerized workloads. Kubernetes v1.36, originally released upstream on April 22, 2026, introduces a suite of improvements, with several critical features now reaching General Availability (GA). Among these is Fine-grained Kubelet API Authorization, which establishes least-privilege access control for the Kubelet API, thereby eliminating the need for overly broad permissions often seen in monitoring and observability use cases. Another significant GA feature is User Namespaces, which substantially improves pod isolation by enabling rootless containers. This enhancement is particularly beneficial for strengthening multi-tenant security within Kubernetes clusters. Furthermore, Mutating Admission Policies have graduated to GA, simplifying policy management by allowing built-in Common Expression Language (CEL)-based policies to replace more complex webhook-based mutations. For environments running advanced workloads, especially in the realm of Artificial Intelligence and Machine Learning (AI/ML), the introduction of Dynamic Resource Allocation (GA) is a crucial update. This feature optimizes the scheduling and management of specialized hardware, such as GPUs, making OKE a more robust platform for AI/ML at scale. These collective improvements aim to bolster security, streamline cluster operations, and provide comprehensive support for the demands of modern applications. OCI customers can begin utilizing Kubernetes v1.36 on OKE by creating new clusters or upgrading existing ones via the Oracle Cloud Console, CLI, API, SDK, and Terraform.
#kubernetes#oci#oke#containers#cloud#security#ai/ml#resource management
Read original source