Release 2.10.0 (2025 Q2)

• Kubernetes version: 1.32.3
• This release is no longer supported. See release policy.

Available versions for transition #

¹ When upgrading from 2.9.x to 2.10.0, perform manual steps if monitoring components were used before the upgrade:

• If the “Local Metrics Collection Module” was enabled in the client cluster, and the “Monitoring Module. Centralized Metrics Collection Component” was enabled in the management cluster, after updating the client cluster, open the operator.victoriametrics.com API group in the management cluster, find the VMRule resource with the prefix system in the namespace of the client cluster and update it spec with the current one update instructions.
• If the “Monitoring module. Centralized metrics collection component” was enabled in the management cluster before the update, after the update add custom resources VMRule, VMAlertmanagerConfig and VMAlert for the correct operation of dashboards in Grafana. In the manifests, specify the namespace of the management cluster, and for VMAlert use the notifier address http://vmalertmanager-shturval-metrics-collector.victoria-metrics:9093.

Infrastructure #

• Added integration with Yandex Cloud provider.
• Added the ability to manage OpenStack provider availability zones.
• Support for OSnova operating system is now available.
• The ability to install a cluster on one Master node (microcluster) is now available.
• The interactive mode of platform installation has been improved: in case of errors or forced termination of the interactive mode, the ability to restart the mode with the last entered data is now available.
• Based on Victoria Metrics, metering is now available in the context of cluster namespaces.
• For the stm (shturval mirror) utility, parameters have been added for: checking the status of the mirror, stopping the mirror and deleting the mirror.
• Added the ability to forcefully delete hosts if they are unavailable in the Shturval V2 provider.
• Added a proxy module for external authentication services. Setting up an authentication connection with an external provider (Keycloak, Blitz) has been simplified.
• Restored Opensearch dashboard in Grafana.
• Added exception of Metadata from collected system workload logs.

Interface changes #

• Added Yandex Cloud provider management GUIs, installation and cluster management GUIs with the Yandex Cloud provider.
• Interfaces for managing multicluster and intracluster tenants have been implemented.
• Added an interface for managing access rights of users/user groups in tenants.
• Added to the cluster dashboard:
  • viewing the metering of cluster namespaces;
  • viewing the validity period of certificates in the cluster;
  • division into groups of Worker nodes.
• Added the ability to force Pod removal.
• ALD Pro has been added to the directory list in the “LDAP Integration” page interface.
• Added the CiliumEgressGatewayPolicy management GUI in the cluster on the “Egress Gateway” page.
• The Ingress Controllers management GUI has been moved to the cluster level.
• In the custom resources interface:
  • the ability to delete a custom resource is now available.
  • added Kubernetes API resource groups (k8s.io, core).
  • added the ability to import a manifest.
• In the Storage Classes interface, the ability to select provider types has been added: Rawfile, CephFS, CephRBD.
• Added transition to the page for viewing an object after its creation/change.
• Added an interface for viewing and changing YAML backup manifests in the cluster and cluster names: Storages, Backups, Schedulers, Restored copies, Repositories.
• Allowed deletion of a Master node without re-creation, if after deletion there remains an odd number of healthy Master nodes.
• Added auto-update of the “Session Manager”, “List of Providers”, “List of Providers” pages in the platform every 5 seconds.
• Added the ability to set Taints (restrictions) on the node and Tolerations (permissions) in the pod specification.
• Added display of MachineDeployment, KubeadmControlPlane states in the configuration of node groups.
• Added search feature:
  • in the service specification and available parameters for service configuration;
  • in YAML object manifests;
  • in the pod logs;
  • In the “Available Charts” interface by chart name.
• Added the ability to manage display and behavior settings in the YAML manifest editor.
• On the “About the system” page and in the cluster update window, links to documentation have been added.
• In the process of creating a cluster at the configuration step of Worker nodes, the ability to duplicate the configuration of the created group and create a duplicate group has been added.
• In the connection configuration settings of the Basis Dynamix provider, the ability to specify the application ID and API key has been added.
• On the “Policy Manager” pagesty" added the ability to import manifests.
• Updated interface of the node view page:
  • network interfaces have been moved to a separate tab;
  • interface space reorganized;
  • added warning about pods being evicted from a node.
• Added the ability to automatically search for users/user groups in the catalog when assigning access rights.
• Added display of configured sections in NCI on the node configuration page.
• Added the ability to enable Swagger and go to Swagger from the “About” page interface.
• Hints have been added for Drain parameters in the Drain node configuration window.
• Renamed the “Network Policies” page to “Cilium Network Policies”.
• In the marketplace of partner products, the ability to connect a pre-configured Luntry repository is available.
• In modal interface windows, the ability to change the width of the window is now available.
• In the Deployment interface, the ability to set the waiting time for deployment in seconds (ProgressDeadlineSeconds) and the number of ReplicaSets that must be saved to be able to rollback (RevisionHistoryLimit) has been added.
• Added a notification about exceeding the license limit if the number of Worker nodes exceeds the number of nodes according to the license.
• Updated YAML manifest display screen.

Bug fixes #

• display port restrictions in network policies.
• displaying network interfaces in the GUI.
• a warning appears about unsaved changes without making changes when moving from the “Routes” page.
• the ability to edit the Openstack provider configuration without checking the connection.
• deleting a host in the Shturval v2 provider in the GUI.
• restart the update.
• creating a cluster with the Basis Dynamix provider.
• removing imageMaximumGCAge from the Kubelet configuration when installing Feature Gate CPUManager.
• readiness of Node Configs of management cluster nodes after installation of the platform.
• ability to select disabled providers in the platform in the cluster creation interface.
• displaying a message when selecting an incorrect platform version with an external cluster profile in the cluster creation interface.
• obtaining the state in the Infra Machine of a cluster node with the Shturval v2 provider, if there is no free host with the specified role.
• displaying potential hosts in the GUI at the step “Configuring groups of Worker nodes” in the process of creating a cluster with the Shturval v2 provider.
• authorization in Grafana when navigating from the cluster GUI.
• installation of the oVirt CSI Module into the cluster after installing a cluster with CSI integration disabled.
• deleting ConfigMaps, Secrets after deleting the key.
• displaying an error in specifying the port in the local repository address when setting up NCI Container Runtime.
• display the title of the modal window when adding a label.
• character restrictions for Pause image addresses.