Release 2.8.0 (2024 Q4)

• Kubernetes version: 1.30.5
• This release is no longer supported. See release policy.

Available versions for transition #

¹ Updating a cluster to version 2.8.0 and 2.8.1 is available only from version 2.7.2. Before starting the update, create the provider-secret secret in the management cluster. This is related to a change in the approach to managing the management cluster. For details, see release 2.7.2.

Infrastructure #

• Updated versions of components: Nginx, NodeLocalDNS, Argo CD, ArgoRollouts, CNI, FluentBit, Grafana, Prometheus, VictoriaMetrics, Kyverno, KubeVIP, Containerd, CertManager, Ceph-csi, OpenSearch, SR-IOV, vSphere, OpenStack, Multus, Whereabouts.
• The default namespace kube-system is assigned to SR-IOV services.
• Added automatic assignment of the network interface name eth0 for providers: vSphere, oVirt, Basic Dynamix, OpenStack.
• Added node certificate management for RHEL.
• Added the ability to configure parameters for the ACME server when installing a management cluster.
• Support for the OCI protocol is provided for connecting repositories to the cluster.
• Added the ability to automatically update from a mirror.
• The interactive mode of installation of the management cluster has been improved.
• The ability to operate the management cluster in stand-alone cluster mode is now available.
• Added support for Astra Linux SE 1.8.
• Community version of Shturval platform has been released. The version does not contain restrictions on the functionality of the platform. The Community version can only be used in an open loop.

Security #

• Kubebench component stopped working. Cluster configuration scanning for compliance with security standards is now performed by Trivy.
• The ability to use the management cluster as a client cluster is now available.
• A logical separation of the functionality of the Platform and clusters is now available. In clusters, management of the management cluster and client clusters is available. The platform includes the necessary functionality for managing infrastructure providers, integration with LDAP, and managing access sets.
• Updated rules for management cluster names: lowercase letters, numbers and hyphens can be used.
• Added a mechanism for viewing pod logs in case of interrupted installation.
• The approach to determining cluster statuses has been changed. Prevented situations where information about the active cluster is retrieved if it has not yet been initialized.
• Improved management of deleting user and user group assignments.
• Added the ability to create a custom dashboard in the Opensearch interface. More details on the page “Add a dashboard in Opensearch” in the “P” sectionstep-by-step instructions."

Interface changes #

• The interface of the “Scan Results” page in the Security section has been updated: information on the results of scanning the cluster configuration is given according to Trivy Compliance reports. More details on the “Scan Results” page of the “Security” section.
• Added English language support for working in the platform GUI. In the user settings, you can select the interface language.
• The interface of the provider configuration page has been updated: setting the InfraMachineTemplate parameters is available on a separate tab.
• Updated interface of the node configuration page (NCI):
  • section settings are divided:
    • trusted and system certificates;
    • kernel arguments and modules;
  • added the ability to view and go to the pages of nodes on which the configuration is planned to be applied (NodeConfig);
  • default values of Kubelet parameters are no longer applied automatically and are included in tooltips.
• The interface of the namespace Loads pages has been updated: the ability to select persistentVolumeClaim for mounting in pod containers has been added to the list of volumes.
• Added the ability to select Security Groups management to the OpenStack provider instance configuration page.
• Updated site page interface:
  • added a tab with a list of pods located on the node;
  • added the ability to track the application of node configurations (NCI) on each node. More details on the “Node Management” page;
  • added the ability to manage cluster network policies. More details on the page “Clusters/Administration/Network Policies”.
• Added the ability to view an extended status model for applying access set assignments for users and groups at the platform, cluster and namespace level.
• The interface of the page for viewing user access rights has been updated: detailed information is available on the assignment of user/group access sets at the platform, cluster and namespace levels.
• The ability to assign access rights in the management cluster and platform has been logically separated: now platform access rights can only be assigned in the “Platform” section.
• The ClusterIssuer and Issuers interface has been improved: the ability to add a root and intermediate certificate of the corresponding certification authority (CA Bundle) is now available.
• The interface of the chart installation pages and the installed service has been updated: the name of the update field has been replaced with management mode, added a hint for the toggle switch to turn on/off the service.
• Added the ability to view documentation from the pages of the Shturval platform GUI.
• Added the ability to switch to viewing a node from pod pages.
• An interface for viewing a report on the requested resources of all client clusters of the platform is now available.
• Added transition to the page for viewing alerts from the cluster dashboard.
• The ability to edit pre-configured access sets on the access set viewing page is blocked.
• Protection has been installed against accidentally pressing the cluster creation button again.
• Updated Platform dashboard interface. Added quantitative data on system and custom access sets, information on connecting a user directory, pools of IP addresses and quantitative characteristics of platform clusters and nodes.

Bug fixes #

• The logic for displaying the “Save” and “Change” buttons has been adjusted: now the buttons are available for clicking if there are changes and if all required fields are filled in.
• Fixed errors during installation in a conditionally closed loop on the Rocky Linux OS.
• Clearing the access binding of a namespace user after deleting a namespace.
• Clear Argo CD role assignment information for a user after deleting access set assignments to a user.
• Filling in fields for automatic integration with LDAP.
• Managing access rights for cluster users.
• Display node group roles in the group configuration interface.
• Stabilized connection to the terminal from the Pod interface.
• Saving the nodeVolumeDetachTimeout value when adding a group of Worker nodes.
• Drain node from the GUI from the GUI.

Known limitations #

• Astra Linux SE version 1.7.5 is not supported in release 2.8.0.
• Creating client clusters is only available with the Platform-admin access set.