Release Stork v2.1.0

New Features

• Added new CRD to support GroupVolumeSnapshots (#222)
  • Examples of the new GroupVolumeSnapshot can be found here
  • The old way to trigger GroupVolumeSnapshots using annotations in the SnapshotVolume spec has been deprecated.

Updates

• The Rule CRD has been updated to conform to other Kubernetes specs (#227)
  • The spec fields has been renamed to rules
  • Examples have been updated here
• Copy storkctl for all architectures into the container (#255)
  • They are placed under /storkctl/<arch> in the container

Bug Fixes

• Update CA certs in container. This helps fix issues when pairing with IKS clusters (#221)
• Don't delete clusterIP when migrating headless service (#231)
• Fixes for tracking pods to perform recovery when running rules (#232 #235)
• Set the namespace annotations from source cluster during migration (#248)
• Don't override pod variable when running a command in the pod for rule execution (#258)
• Portworx:
  • Use volumeID instead of volume name after creating snapshot (#239)

Release Stork v2.0.1

Bug Fixes

• Portworx: Fix regex for version matching when checking feature compatibility.

Release Stork v2.0.0

New Features

• Support for pairing clupters and migrating stateful applications between clusters
  • Requires a storage driver which supports migrating PVCs between 2 paired clusters.
  • Added support in Portworx driver for Migration
• storkctl to help manage custom resource
  • Prints out more helpful information for CRDs and helps create them with options from command line

Docs are being re-organised, will add link here with information for the new features when done.

Release Stork v1.2.1

Bug Fixes

• When logging pod info check for owner pointer before dereferencing (#204)
• Portworx: Also check for CSI provisioner name for ownership (#214)

v1.3.0-beta1

Update version to 1.3.0-beta

Release Stork v1.2.0

New Features

• Support for taking application consistent snapshots using pre and post snapshot hooks. Documentation on creating rules and using them in snapshots can be found here

Release Stork v1.1.4

Bug Fixes

• Extender: Match driver and scheduler nodes using internal IP if names don't match (#142)
• Health Monitor: Update logic to match nodes similar to Extender (#143)

Release Stork v1.1.3

Bug Fixes

• Extender: Also check node name to match driver nodes (#112)
  • This helps deal with clusters where the IP is set as the hostname
• Portworx: Remove dependency on source PVC when restoring from cloudsnap (#120)

Release Stork v1.1.2

Bug Fixes

• Portworx Driver
  • Fix restore from group snapshots (#106)
  • Query node cache with id and hostname

Release Stork v1.1.1

Improvements

• Allow snapshots to be restored to different namespaces (#71)
  • When creating snapshots users need to provide comma seperated regexes
    with "stork/snapshot-restore-namespaces" annotaion to specify which
    namespaces the snapshot can be restored to
  • When creating PVC from snapshots, if a snapshot exists in another
    namespace, the snapshot namespace should be specified with
    "stork/snapshot-source-namespace" annotation

Bug fixes

• Use cache for node infomation to prevent timeouts during scheduling in large clusters (#99)