Release 2.9.0 (2024 Q4)

Infrastructure #

• Improved platform interactive installation mode: cluster configuration type selection parameter (ha/single) removed.
• Platform installation from a workstation with Windows OS is now available. Installation still runs only on Linux hosts.
• Local metrics collection component (Prometheus) replaced with VictoriaMetrics Agent (shturval-metrics-collector).
• Local logs collection component (Fluentbit) replaced with Vector (shturval-log-collector).
• Mirror certificate rotation (ACME) and wildcard certificate installation support.
• Automatic kubectl-shturval plugin configuration.

Interface changes #

• Improved Shturval v2 provider and node group management GUIs.
• In workload objects:
  • Security Context GUIs added;
  • request and limit unit selection added.
• Improved directory search in access assignment GUI:
  • search accelerated;
  • partial match search available;
  • search now available only by unique group or user name.
• “Available Charts” page GUI: added chart installation from repositories connected via “oci” protocol.
• In “Node Management” GUI:
  • requested and healthy node counters added; node list scrolling in groups implemented.
  • “InfraMachineTemplates” tab added for infrastructure template management.
  • “Machinehealthcheck” GUI updated in “Node group configuration” block. Drain launch parameters added to node view GUI.
• StorageClass GUI: custom parameter addition added.
• Secrets, Configmaps display GUI updated.
• velero namespace addition excluded when creating backup from GUI.
• “Backup” and “Alerts” section pages: object view and edit including via YAML added.
• Ingress status display in namespaces improved.
• Object create, modify, and delete notification text corrected.
• “Repositories” page: connected repository status display added.
• Namespace list page: pagination added.
• ManagedFields hidden in object YAML manifest view and edit GUI.

Bug fixes: #

• oVirt-CSI installation.
• NCI application on Master nodes in vSphere provider cluster.
• network policy rule creation when selecting services.
• UserRoles deletion when deleting cluster.
• kubeconfig download from GUI.
• localization (translation) in GUI.
• object access rights in GUI.
• host state detection in shturval v2 provider.
• Master node count change in shturval v2 provider cluster.
• backup deletion after storage deletion.
• restore from backups (velero namespace exclusion from backups).
• root administrator check during platform authentication.
• deployment pause from GUI.

Known issues #

• Node drain may ignore selector values.

Breaking changes #

• Log aggregation principle changed according to types (auditd, journald, shturvald, logs, user-logs, events, kube-audit). New Index Patterns must be created in OpenSearch.