* Bug fixes

  * Fixes issue with Flexvolume, introduced in 1.6.0, where drivers without an attacher would fail (node indefinitely waiting for attach). A driver API addition is introduced: drivers that don't implement attach should return attach: false on init. ([#47503](https://github.com/kubernetes/kubernetes/pull/47503), [@chakri-nelluri](https://github.com/chakri-nelluri))

### Deprecation

- Added support for vSphere in-tree volumes migration to vSphere CSI driver. The in-tree vSphere Volume plugin will be deprecated and removed in a future release.
  

#### **Other**

*   Pod Security Policy can now manage access to specific FlexVolume drivers ([#53179](https://github.com/kubernetes/kubernetes/pull/53179),[ @wanghaoran1988](https://github.com/wanghaoran1988))

	"github.com/minio/pkg/v3/mimedb"
	ftp "goftp.io/server/v2"
)

var _ ftp.Driver = &ftpDriver{}

// ftpDriver implements ftpDriver to store files in minio
type ftpDriver struct {
	endpoint string
}

// NewFTPDriver implements ftp.Driver interface
func NewFTPDriver() ftp.Driver {
	return &ftpDriver{endpoint: fmt.Sprintf("127.0.0.1:%s", globalMinioPort)}
}

### Action Required

* ACTION REQUIRED: Removes defaulting of CSI file system type to ext4. All the production drivers listed under https://kubernetes-csi.github.io/docs/drivers.html were inspected and should not be impacted after this change. If you are using a driver not in that list, please test the drivers on an updated test cluster first. ``` ([#65499](https://github.com/kubernetes/kubernetes/pull/65499), [@krunaljain](https://github.com/krunaljain))

			// Bucket or prefix/directory not found
			hr.Before.Drives[i] = madmin.HealDriveInfo{Endpoint: drive, State: madmin.DriveStateMissing}
			hr.After.Drives[i] = madmin.HealDriveInfo{Endpoint: drive, State: madmin.DriveStateMissing}
		default:
			hr.Before.Drives[i] = madmin.HealDriveInfo{Endpoint: drive, State: madmin.DriveStateCorrupt}
			hr.After.Drives[i] = madmin.HealDriveInfo{Endpoint: drive, State: madmin.DriveStateCorrupt}
		}
	}

	}

	// Fetch all the drive info status.
	beforeDrives := formatsToDrivesInfo(s.endpoints.Endpoints, formats, sErrs)

	res.After.Drives = make([]madmin.HealDriveInfo, len(beforeDrives))
	res.Before.Drives = make([]madmin.HealDriveInfo, len(beforeDrives))
	// Copy "after" drive state too from before.
	for k, v := range beforeDrives {
		res.Before.Drives[k] = v
		res.After.Drives[k] = v
	}

drives.  Therefore, the number of drives you present must be a multiple of one of these numbers.  Each object is written to a single erasure-coding set.

Minio uses the largest possible EC set size which divides into the number of drives given. For example, *18 drives* are configured as *2 sets of 9 drives*, and *24 drives* are configured as *2 sets of 12 drives*.  This is true for scenarios when running MinIO as a standalone erasure coded deployment. In [distributed setup however node (affinity)...

* Docker versions 1.10.3, 1.11.2, 1.12.6 have been validated
  * Docker version 1.12.6 known issues
    * overlay2 driver not fully supported
    * live-restore not fully supported
    * no shared pid namespace support
  * Docker version 1.11.2 known issues
    * Kernel crash with Aufs storage driver on Debian Jessie ([#27885](https://github.com/kubernetes/kubernetes/issues/27885))

- **MinIO creates erasure-coding sets of _2_ to _16_ drives per set.  The number of drives you provide in total must be a multiple of one of those numbers.**
- **MinIO chooses the largest EC set size which divides into the total number of drives or total number of nodes given - making sure to keep the uniform distribution i.e each node participates equal number of drives per set**.