|
| 1 | +# v0.6.0 - Changelog Since v0.5.0 |
| 2 | + |
| 3 | +## Breaking Changes |
| 4 | + |
| 5 | +- Some of the API objects in the deployment specs have changed names/labels/namespaces, please tear down old driver before deploying this version to avoid orphaning old objects. You will also no longer see the driver in the `default` namespace. |
| 6 | +- Some error codes have been changed, please see below for details if you rely on specific error codes of the driver |
| 7 | + |
| 8 | +## New Features |
| 9 | + |
| 10 | +- Add support for Raw Block devices. ([#283](https://github.com/kubernetes-sigs/gcp-compute-persistent-disk-csi-driver/pull/283), [@davidz627](https://github.com/davidz627)) |
| 11 | +- Operations in the node driver are now parallelized, except those involving a volume already being operated on now return an error. ([#303](https://github.com/kubernetes-sigs/gcp-compute-persistent-disk-csi-driver/pull/303), [@hantaowang](https://github.com/hantaowang)) |
| 12 | +- Adds support for ControllerExpandVolume and NodeExpandVolume ([#317](https://github.com/kubernetes-sigs/gcp-compute-persistent-disk-csi-driver/pull/317), [@davidz627](https://github.com/davidz627)) |
| 13 | +- Operations in the controller driver on a volume already being operated on now return an error. ([#316](https://github.com/kubernetes-sigs/gcp-compute-persistent-disk-csi-driver/pull/316), [@hantaowang](https://github.com/hantaowang)) |
| 14 | +- Picking up support for inline volume migration and some fixes for backward compatible access modes for migration ([#324](https://github.com/kubernetes-sigs/gcp-compute-persistent-disk-csi-driver/pull/324), [@davidz627](https://github.com/davidz627)) |
| 15 | + |
| 16 | + |
| 17 | +## Bug Fixes |
| 18 | + |
| 19 | +- Reduces node attach limits by 1 since the node boot disk is considered an attachable disk ([#361](https://github.com/kubernetes-sigs/gcp-compute-persistent-disk-csi-driver/pull/361), [@davidz627](https://github.com/davidz627)) |
| 20 | +- Fixed a bug that causes disks in the same zone/region to be provisioned serially ([#344](https://github.com/kubernetes-sigs/gcp-compute-persistent-disk-csi-driver/pull/344), [@hantaowang](https://github.com/hantaowang)) |
| 21 | +- Remove cross validation of access modes, multiple access modes can be specified that represent all the capabilities of the volume ([#289](https://github.com/kubernetes-sigs/gcp-compute-persistent-disk-csi-driver/pull/289), [@davidz627](https://github.com/davidz627)) |
| 22 | +- Driver should check socket parent directory before trying to bind it ([#339](https://github.com/kubernetes-sigs/gcp-compute-persistent-disk-csi-driver/pull/339), [@zhucan](https://github.com/zhucan)) |
| 23 | +- Updated CSI Attacher to stop ignoring errors from ControllerUnpublish ([#378](https://github.com/kubernetes-sigs/gcp-compute-persistent-disk-csi-driver/pull/378), [@davidz627](https://github.com/davidz627)) |
| 24 | +- CreateVolume will now fail with NOT_FOUND error when VolumeContentSource SnapshotId does not refer to a snapshot that can be found ([#312](https://github.com/kubernetes-sigs/gcp-compute-persistent-disk-csi-driver/pull/312), [@davidz627](https://github.com/davidz627)) |
| 25 | +- ControllerUnpublishVolume now returns success when the Node is GCE API NotFound. |
| 26 | +Invalid format VolumeID is now GRPC InvalidArgument error instead of GRPC NotFound. |
| 27 | +Underspecified disks not found in any zone now return GRPC NotFound. ([#368](https://github.com/kubernetes-sigs/gcp-compute-persistent-disk-csi-driver/pull/368), [@davidz627](https://github.com/davidz627)) |
| 28 | + |
| 29 | + |
| 30 | +## Other Notable Changes |
| 31 | + |
| 32 | +- Deployment spec updates: |
| 33 | +The deployment is no longer in namespace `default` |
| 34 | +Changed "app" label key to "k8s-app" |
| 35 | +csi-snapshotter version has been changed to v1.2.0-gke.0 |
| 36 | +The resizer role binding has been renamed to "csi-controller-resizer-binding" |
| 37 | +Removed driver-registrar role. ([#364](https://github.com/kubernetes-sigs/gcp-compute-persistent-disk-csi-driver/pull/364), [@verult](https://github.com/verult)) |
| 38 | +- Updating the following image versions in stable deployment specs: |
| 39 | +gcp-compute-persistent-disk-csi-driver: v0.6.0-gke.0 |
| 40 | +csi-provisioner: v1.4.0-gke.0 |
| 41 | +csi-attacher: v2.0.0-gke.0 |
| 42 | +csi-node-driver-registrar: v1.2.0-gke.0 ([#400](https://github.com/kubernetes-sigs/gcp-compute-persistent-disk-csi-driver/pull/400), [@verult](https://github.com/verult)) |
0 commit comments