Release 2.9.0 (2024 Q4)

Available versions for transition

This version has been decommissioned. If necessary, use the patch version.

Infrastructure

  • The interactive platform installation mode has been improved: the option to select the type of cluster configuration (ha/single) has been eliminated.
  • The ability to launch the platform installation from a working machine with Windows installed is now available. Installation is still performed only on hosts with Linux OS.
  • The local metrics collection component (Prometheus) has been replaced with VictoriaMetrics Agent (shturval-metrics-collector).
  • The local log collection component (Fluentbit) has been replaced with Vector (shturval-log-collector).
  • Added support for rotation (ACME) and installation of wildcard certificates for the mirror.
  • Added automatic configuration of the kubectl-shturval plugin.

Interface changes

  • Improved interfaces for managing the provider and groups of nodes with the Shturval v2 provider.
  • In Loads objects:
    • added Security Context interfaces;
    • added the ability to select units of measurement for requests and limits.
  • Improved catalog search in the GUI for assigning access rights:
    • search is accelerated;
    • the ability to search by partial match is now available;
    • the search option is now available only by unique group or user name.
  • The ability to install charts from repositories connected via the “oci” protocol has been added to the interface of the “Available Charts” page.
  • In the “Node Management” interface:
    • added counters of requested and healthy nodes, implemented the ability to scroll the list of nodes in groups.
    • added the “InfraMachineTemplates” tab for managing infrastructure templates.
    • updated the “Machinehealthcheck” interface in the “Node group configuration” block. Parameters for launching Drain have been added to the node viewing interface.
  • Added the ability to add custom parameters to the StorageClass interface.
  • Updated interface for displaying Secrets, Configmaps.
  • The possibility of adding the velero namespace when creating a backup from the GUI is excluded.
  • On the pages of the “Backup” and “Alerts” sections, the ability to view and edit objects has been added, incl. using YAML.
  • Improved display of Ingress status in namespaces.
  • The text of notifications about the creation, modification or deletion of objects has been adjusted.
  • Added connection status display to the “Repositories” page for unlocked repositories.
  • Pagination has been added to the page with a list of namespaces.
  • The ManagedFields display is hidden in the GUI for viewing and editing YAML object manifests.

Bug fixes

  • oVirt-CSI installations.
  • application of NCI on Master nodes in a cluster with the vSphere provider.
  • creating network policy rules when selecting services.
  • deleting UserRoles when deleting a cluster.
  • downloading kubeconfig of the cluster from the GUI.
  • localization (translation) in the GUI.
  • access rights to objects in the GUI.
  • determining the state of hosts in the shturval v2 provider.
  • changes in the number of Master nodes in a cluster with the shturval v2 provider.
  • deleting backup copies after deleting the storage.
  • restoration from backup copies (the possibility of the velero namespace getting into backup copies is excluded).
  • checking the root administrator when authenticating to the platform.
  • pausing deployment from the GUI.

Known issues

  • Drain node can ignore selector values.

Breaking changes

  • The principle of log aggregation has been changed in accordance with types (auditd, journald, shturvald, logs, user-logs, events, kube-audit). You need to create new Index Patterns in OpenSearch.
← Release 2.9.1 (2025 Q1)
Release 2.8.1 (2024 Q4) →