openstack-cinder-csi 2.36.4 - Helm Chart Release for OpenStack CSI Driver

The cloud-provider-openstack repository released openstack-cinder-csi-2.36.4 on September 7, 2026. This release updates the Helm chart deployment manifests for the OpenStack Cinder Container Storage Interface driver. It is a stable patch release rather than a prerelease, targeted at production Kubernetes clusters requiring persistent block storage integration on OpenStack infrastructure.

The full release notes and downloads are on the GitHub release page.

The openstack-cinder-csi-2.36.4 tag updates the primary Helm chart for deploying the OpenStack Cinder CSI driver. Platform operators use this chart to manage the life cycle of CSI controller deployments and node agent daemonsets across Kubernetes clusters. The update aligns chart templates with the 2.36 release line of cloud-provider-openstack, ensuring compatible API group versions and container image defaults.

Helm chart updates in this release streamline installation for self hosted Kubernetes clusters operating on OpenStack clouds. The chart encapsulates RBAC definitions, service accounts, storage class parameters, and secret bindings needed for volume provisioning. Operators running automated helm diff pipelines will observe updated default values and chart metadata matching the patch release.

The Cinder CSI driver enables dynamic provisioning of OpenStack Cinder volume resources inside Kubernetes. Data engineering workloads rely on Cinder block storage for persistent state, including stateful query engines, distributed stream buffers, and time series databases. The CSI driver handles volume attachment, detachment, expansion, and snapshot operations directly through the OpenStack block storage API.

Using out of tree volume drivers like openstack-cinder-csi isolates storage controller logic from the core kube-controller-manager binary. This separation allows storage components to be upgraded independently of the main Kubernetes control plane. Deploying the chart from openstack-cinder-csi-2.36.4 ensures that storage operations execute against tested OpenStack API client routines.

In production environments, the CSI driver operates across two distinct workload topologies. The CSI controller deployment runs as a replicated set on control plane nodes to handle volume creation and delete requests. Simultaneously, the CSI node daemonset runs on every worker node to handle local device mounting, filesystem formatting, and path binding.

Updating to openstack-cinder-csi-2.36.4 helps maintain operational parity across multi zone OpenStack deployments. When node agents attach Cinder volumes, topology constraints ensure that volumes match the availability zone of the scheduled pod. Platform teams should verify that chart values match cloud credentials stored in local cluster secrets before applying updates across production environments.

Upgrading an existing Cinder CSI deployment requires executing a standard Helm upgrade command against the release repository. Operators should review custom values files for deprecated parameter names before deploying. Because node daemonsets manage active volume mounts, rolling out node updates should be coordinated with maintenance windows or node drain procedures to prevent downtime for running storage workloads.

Before running helm upgrade, verify that cloud-config secrets are present in the target namespace and contain valid OpenStack authentication details. Testing the upgrade in a non production cluster allows validation of dynamic volume creation, volume attachment, and snapshot deletion routines before production rollout.