Release 2.7.0 (2024 Q3)

Infrastructure

  • Updated component versions: ArgoCD, ArgoRollouts, OpenStack, CertManager, CNI Cilium, Grafana, KubeVIP, Nginx, NodeLocalDNS, Prometheus, Whereabouts (SR-IOV), VictoriaMetrics, vSphere, Velero, RawFileProvisioner (updated images: csi-provisioner, csi-node-driver-registrar, csi-resizer), KubeBench, Trivy, Containerd.
  • Implemented a utility for deploying and updating a mirror in Nexus (including customer Nexus). See installation instructions for closed and semi-closed environments for details.
  • Added the ability to protect nodes from automatic node actions (drain, reboot). Protection management is available to the platform administrator from the console. See How to protect nodes from automatic restart page for details.
  • Added HealthChecks for services installed in the management and client clusters.
  • Updated AlertManager configuration validation for stable operation when invalid data is detected. AlertManager changes cannot be applied with invalid configuration.
  • Discontinued support for the internal Shturval Manual platform provider. Implemented a mechanism blocking update to 2.7.0 and above when a cluster deployed with Shturval Manual provider exists.
  • Updated the list of critically important services for cluster operation. Services for virtual IP management, external connections (Ingress), and authentication (authn) now have guaranteed scheduling space in the cluster. Critical services cannot be disabled or deleted. Critical service pods cannot be evicted during workload distribution across nodes.
  • Accelerated Shturval v2 provider host processing.
  • Added RED OS 8 operating system support.
  • Implemented IP address pool management and the ability to assign pool addresses to the cluster API server and Ingress when deploying clusters with vSphere, oVirt, and Shturval V2 providers.
  • Added the ability to manage ClusterAPI Machine, KubeadmControlPlane, InfraMachineTemplate, MachineHealthCheck configuration from the GUI. The GUI is available on the Node Management page.
  • Updated Master node configuration method when creating a cluster. Cluster availability type setting (single, ha) replaced with Master node count request.

Security

  • Added the ability to authenticate users via kubectl without using the platform GUI.
  • Implemented access set configuration management by the platform administrator for the management cluster, client clusters, and namespaces.
  • Implemented user and group search in the directory (AD) when assigning access rights to users/groups. Directory search is available when LDAP integration is successfully configured. User and group search GUI is available on Access Assignment pages in the management and client clusters.
  • Implemented Worker node group management for clusters.

Interface changes

  • Implemented YAML manifest view and edit GUI for HPA in Deployment and network policies.
  • Updated network policy editor GUI in client clusters. Added namespace selection from list; DNS proxy enabled by default when traffic to Kubernetes DNS is allowed.
  • Implemented backup and restore and custom resources management GUIs in management and client cluster namespaces.
  • Added namespace and cluster-scoped resource management in backup and restore GUI.
  • Updated cluster GUI: renamed “Services and repositories” section and pages.
  • Added the ability to set Seccomp profile in workloads of client cluster and management cluster namespaces.
  • Updated installed services view page GUI: service labels and annotations moved to separate tabs; added service configuration parameters (values) view.
  • Updated Certificates block GUI in NCI: added hints and restrictions on certificate data entry method.
  • Improved Egress tab GUI on the namespace page: added informational notification when a network interface name is specified.
  • Updated cluster GUI: added cluster deletion lock and informational message when infrastructure reconciliation is paused.
  • Implemented GUI for managing Worker node groups (Default, Infra, and custom) when creating clusters with vSphere, oVirt, Basic Dynamix, OpenStack, Shturval v2 providers.
  • Implemented product marketplace GUI. Marketplace is available in the “Services and repositories” section.
  • Management cluster GUI moved to the clusters section. Additionally implemented in the management cluster:
    • security events and audit logs retrieval;
    • security policy (Kyverno) management;
    • cluster configuration scanning (KubeBench);
    • container image vulnerability scanning (Trivy);
    • custom resources management;
    • network policy management.
  • Added user interface settings management. The feature is available on the User Settings page.
  • Improved triggered security policy report display in namespaces and namespace workloads.

Fixed

  • Fixed node cleanup handling in the node configuration management module (STD).
  • Fixed automatic redirect after session termination. Re-authentication redirects to the page where the session was interrupted.
  • Added information to authentication service (authn) logs on unsuccessful user authentication.
  • Added input validation for Shturval v2 provider host IP addresses when configuring from console.
  • Fixed cyclic shturvald (STD) restart error when manually creating a directory in /etc/containerd/certs.d/. In such cases, nodes in clusters no longer receive “SchedulindDisabled” (Unschedulable) status.
  • Fixed reconciliation and deletion button display errors on cluster dashboard when description block information is missing.
  • Fixed security events retrieval error when navigating to “Security Events” in client cluster GUI.
  • Fixed ArgoCD navigation availability display error when the service is absent.
  • Fixed incorrect access rights assignment error when selecting “Cluster-admin-restricted” role.
  • Fixed namespace object retrieval error when switching between cluster namespaces.
  • Fixed YAML manifest format validation unavailability in GUI after returning valid structure.
  • Fixed missing URL in metrics visualization module specification.
  • Fixed external URL data display error in namespace Ingresses GUI.
  • Pod selector editing in Deployment GUI is prohibited.
  • Fixed LDAP connection error with TLS connection security check.
  • Fixed connection setup check error in automatic mode.
  • Fixed Control Plane node data display error in Shturval v2 provider cluster after cluster update.
  • Fixed Events retrieval error for namespace resources.

Known limitations:

  • Client cluster creation is available only with Platform-admin access set.
  • There may be access issues to the “Scan Results” page.
  • Kubelet section excluded from node configuration GUI. Will be returned in release 2.7.1.
← Release 2.7.1 (2024 Q3)
Release 2.6.1 (2024 Q3) →