Skip to content

Latest commit

 

History

History
119 lines (75 loc) · 5.96 KB

CHANGELOG.md

File metadata and controls

119 lines (75 loc) · 5.96 KB
Raw

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[Unreleased]

[0.2.3] - 2022-12-12

Fixed

  • Hide noisy error message for PVs pointing to non-existent SC behind -v=5 or more

[0.2.2] - 2022-12-12

Fixed

  • Regression in v0.2.0 - it was correct to use non-rate limited functions in certain scenarios, rate limit should only be used for retry after errors
  • Retry in case of errors was broken - it was using Forget instead of Done which was just resetting the rate limit counter but not actually removing previous item from the queue
  • No longer watch for SC - lister is already using efficient caching API, so it is cheap to get a SC by name from the Index during PV reconciliation; that removes some race condition when PV was reconciled and forgotten about before Releaser learned that SC is annotated
  • Do not re-queue an object that has the same version
  • There was race conditions when claimRef was already removed from the PV but Kubernetes didn't yet marked it as Available, re-queuing such a PV was resulting in duplicative Released events - now Released verifies that the claimRef is not already nil before trying to release the volume
  • De-queue objects that are deleted from the cluster

[0.2.1] - 2022-12-11

Fixed

  • Regression in v0.2.0 caused Releaser to panic on concurrent managed SC read

[0.2.0] - 2022-12-11

BREAKING CHANGES

Prior to this change, one of the features of Releaser was to "Automatically associates Releaser with PVs claimed by PVCs that were created by Provisioner with the same --controller-id". From the README.md prior to this change:

PV Releaser Controller

For Releaser to be able to make PVs claimed by Provisioner Available after PVC is gone - Releaser and Provisioner must share the same Controller ID.

Associate

Once Released - PVs doesn't have any indication that they were once associated with a PVC that had association with this Controller ID. To establish this relation - we must catch it while PVC still exists and mark it with our label. If Releaser was down the whole time PVC existed - PV could never be associated making it now orphaned and it will stay as Released - Releaser can't know it have to make it Available.

Releaser listens for PV creations/updates. The following conditions must be met for a PV to be associated with a Releaser:

  • PV doesn't already have metadata.labels."reclaimable-pv-releaser.kubernetes.io/managed-by" association.
  • spec.claimRef must refer to a PVC that either has metadata.labels."dynamic-pvc-provisioner.kubernetes.io/managed-by" or reclaimable-pv-releaser.kubernetes.io/managed-by set to this Controller ID. If both labels are set - both should point to this Controller ID.
  • --disable-automatic-association must be false.

To establish association Releaser will set itself to metadata.labels."reclaimable-pv-releaser.kubernetes.io/managed-by" on this PV.

As disclaimed - that approach was error prone. It was fine most of the time, but if Releaser was down for any noticeable duration of time - it was resulting in PVs piling up in Released state, and as the PVC was long gone by then - PVs would remain in that state forever, until manually cleared up.

This mechanism of association through PVC was removed in this release and replaced with a simple Storage Class annotation. In order for Releaser to turn Released PV as Available - its Storage Class must be annotated with metadata.annotations."reclaimable-pv-releaser.kubernetes.io/controller-id" pointing at the -controller-id of this Releaser. It can now retro-actively release PVs on startup that it never received events about. As a side effect - -controller-id of Provisioner and Releaser doesn't have to match anymore. This unfortunately requires that you use dedicated Storage Class for PVs that must be reclaimable, but that is a fair price to pay if the alternative is unreliable and error prone and might result in expensive storage bills.

You must use Helm charts version v0.1.0 or above as RBAC is changed in this release to allow read-only access to Storage Classes.

Changed

  • Old PV association via PVC mechanism was removed
  • -disable-automatic-association option on Releaser was removed
  • PVs will only be released now if their Storage Class annotated with metadata.annotations."reclaimable-pv-releaser.kubernetes.io/controller-id" pointing at the -controller-id of this Releaser

[0.1.1] - 2022-12-11

Fixed

  • Fixed re-queuing code, previously duplicate items in the queue on updates
  • We was not using graceful rate-limit aware code in some places
  • Fix leader elect code, previously it would not release the lease after a graceful stop
  • Improve basic example with more validation steps
  • Start publishing to GitHub Docker Registry ghcr.io

[0.1.0] - 2022-05-14

Changed

  • Updated with Go 1.18 and K8s dependencies 1.24.0.
  • Fixed NP panic in releaser when labels were not present on the PV.
  • Bumping to v0.1 - I've been using it for a long while now in prod, and it seems to be working fine.

[0.1.0-alpha1] - 2022-04-26

Changed

  • Updated with Go 1.17 and K8s dependencies 1.23.6.

[0.0.4] - 2021-06-20

Added

  • Releaser now checks for both dynamic-pvc-provisioner.kubernetes.io/managed-by and reclaimable-pv-releaser.kubernetes.io/managed-by labels on the PVC for association making it more independent for use cases where Provisioner is not used at all.

[0.0.3] - 2021-06-20

Fixed

  • Do not remove PV association if -disable-automatic-association was true - we didn't set it in the first place

[0.0.2] - 2021-06-19

Fixed

Added

  • Helm charts

[0.0.1] - 2021-06-19

Added

  • Initial implementation