Snapscheduler Versions Save

Added

• Ability to set a priorityClassName for the operator via Helm chart
• Ability to set pod labels and annotations for the operator via Helm chart

Fixed

• Improve cronspec validation to allow "slash" notation

Added

• Ability to specify a container hash instead of just a tag when deploying via Helm chart

Changed

• :warning: Helm chart now directly manages CRD installation/upgrade. If upgrading from 3.1.0 or earlier, manual steps are required:

  Error: UPGRADE FAILED: rendered manifests contain a resource that already
  exists. Unable to continue with update: CustomResourceDefinition
  "snapshotschedules.snapscheduler.backube" in namespace "" exists and cannot be
  imported into the current release: invalid ownership metadata; label
  validation error: missing key "app.kubernetes.io/managed-by": must be set to
  "Helm"; annotation validation error: missing key "meta.helm.sh/release-name":
  must be set to "snapscheduler"; annotation validation error: missing key
  "meta.helm.sh/release-namespace": must be set to "backube-snapscheduler"
  

  The above error can be fixed by adding the required labels and annotations as mentioned in the error message:

  $ kubectl label crd/snapshotschedules.snapscheduler.backube app.kubernetes.io/managed-by=Helm
  customresourcedefinition.apiextensions.k8s.io/snapshotschedules.snapscheduler.backube labeled
  
  $ kubectl annotate crd/snapshotschedules.snapscheduler.backube meta.helm.sh/release-name=snapscheduler
  customresourcedefinition.apiextensions.k8s.io/snapshotschedules.snapscheduler.backube annotated
  
  $ kubectl annotate crd/snapshotschedules.snapscheduler.backube meta.helm.sh/release-namespace=backube-snapscheduler
  customresourcedefinition.apiextensions.k8s.io/snapshotschedules.snapscheduler.backube annotated
  

• Made CRD validation of cronspec more permissive

• Upgrade quay.io/brancz/kube-rbac-proxy to v0.13.1

• Upgrade operator-sdk to 1.23.0

Added

• Ability to control TopologySpreadConstraints via Helm chart

Changed

• Upgrade quay.io/brancz/kube-rbac-proxy to v0.13.0
• Upgrade operator-sdk to 1.22.0

Changed

• Snapshot objects are now accessed via snapshot.storage.k8s.io/v1 API version
• Upgrade operator-sdk to 1.18

Removed

• Removed support for Kubernetes versions < 1.20

Added

• Ability to configure resource requests for RBAC proxy container when deploying via Helm chart.
• Ability to configure container image used for kube-rbac-proxy

Changed

• Build w/ Go 1.17
• Upgrade kube-rbac-proxy image to 0.11.0
• Upgrade operator-sdk to 1.15

Changed

• Updated project scaffolding to operator-sdk 1.10
• Moved CRD to apiextensions.k8s.io/v1
• Added default host anti-affinity for the operator replicas
• Updated Helm Chart manifests to more closely match OSDK scaffolding

Removed

• Removed support for Kubernetes versions < 1.17
• Removed support for snapshot.storage.k8s.io/v1alpha1 snapshot version
• Removed node selector labels targeting beta.kubernetes.io/arch and beta.kubernetes.io/os

Changed

• Switched the operator base container to distroless

Fixed

• Metrics weren't accessible from the snapscheduler-metrics Service

Fixed

• Fix crash when snapshotTemplate is not defined in schedule

Added

• Support Kubernetes 1.17 and snapshot.storage.k8s.io/v1beta1 snapshot version

Added

• Crontab-based schedule CR to take snapshots of CSI-based persistent volumes
• Label selectors to control which PVCs are selected for snapshotting
• Retention policies based on snapshot age or count