1. Docs
  2. Changes and releases
  3. Fleet Management

Fleet Management

  • Fleet Management rook-operator-app v2.0.0

    Added

    • PriorityClasses for osds/mons (#42).
    • Subchart to manage the cluster resources (uses upstream chart) (#40).
    • Make the Ceph version configurable from the main chart’s values (#40).
    • Extended UPDATING.md to reflect the additional subchart (#40).

    Changed

    • Disable CephFS CSI driver by default (#43).
  • Fleet Management vertical-pod-autoscaler-app v1.2.0

    Changed

    • Migrate to configuration management.
    • Update architect-orb to v4.0.0.
  • Fleet Management kubectl-gs v1.35.0

    Changed

    • Apply only v1alpha3 CRs on provider AWS.

    Added

    • Add CRs to create bastion host, when creating a CAPI cluster via template cluster command.
    • Add configuration to allow SSH for Giant Swarm employees when creating CAPI cluster. Applies to template cluster and template nodepool commands.
    • Update template version for CAPA cluster and nodepool templating to version 0.6.8
  • Fleet Management apiextensions v3.32.0

    Added

    • Add support for new matcher types in silence.monitoring.giantswarm.io/v1alpha1 CRD
  • Fleet Management rook-operator-app v1.1.0

    Changed

    • Updated upstream chart to Rook v1.7.0 (#34)
    • Restrict the app to KVM clusters only (#35)
  • Fleet Management apiextensions v3.31.0

    Added

    • Add required field clusterName and label cluster.x-k8s.io/cluster-name for v1alpha3 CAPI CR’s.
  • Fleet Management rook-operator-app v1.0.0

    Added

    • Add alerting rules for managed Ceph clusters (#31).
    • Add servicemonitor to enable monitoring on MCs (#30).

    Changed

    • Promoted rook-operator-app to the main giantswarm-catalog (#32).
  • Fleet Management kubectl-gs v1.34.0

    Throughout our UIs and documentation we are aligning our terminology regarding cluster and node pool details, to use consistent terminology matching our Management API. The unique, immutable identifier that was formerly called ID in our user interfaces, is now called the name. The user-friendly, changeable description of the cluster’s and node pool’s purpose was called name in our UIs and is now called the description. Warning: This terminology change results in a breaking change in the template cluster command, as the purpose of the flag --name has changed. Also several flags in other commands have been deprecated. If you are upgrading from an earlier releases, apply these changes to migrate any scripts:

    • When using template cluster, replace --name with --description to set the user-friendly cluster description, and replace --cluster-id with --name to set the cluster’s unique identifier.
    • When using template nodepool, replace --cluster-id with --cluster-name.

    Added

    • template app: Added the --namespace-annotations and --namespace-labels flags to allow users to specify the namespaceConfig of the generated App manifest. Read App CR’s target namespace configuration for more information.

    Changed

    • get clusters:
      • The output table header ID has been renamed to NAME.
    • get nodepools:
      • The --cluster-id flag is now deprecated, replaced with --cluster-name.
      • Output column headers have been renamed from ID to NAME and from CLUSTER ID to CLUSTER NAME.
    • template cluster:
      • Deprecated the --cluster-id flag.
      • Breaking: the --name flag changed purpose to set the cluster’s unique identifier.
      • The --description flag has been added to set the user-friendly description.
    • template nodepool:
      • Deprecated the --cluster-id flag, added the --cluster-name flag as a replacement.
      • Deprecated the --nodepool-name flag, add the --description flag as a replacement.
  • Fleet Management apiextensions v3.30.0

    Added

    • Add Chart.Description, Chart.Keywords and Chart.UpstreamChartVersion metadata to AppCatalogEntry CRD.
    • Add documentation of customer facing Azure annotations.
  • Fleet Management rook-operator-app v0.3.3

    Changed

    • Revert ‘Ensure the app is deployed to the correct namespace on MCs’.