ceph-csi/internal/cephfs/controllerserver.go

1054 lines
34 KiB
Go
Raw Normal View History

2018-03-05 11:59:47 +00:00
/*
Copyright 2018 The Ceph-CSI Authors.
2018-03-05 11:59:47 +00:00
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
*/
package cephfs
import (
"context"
"errors"
"fmt"
"github.com/ceph/ceph-csi/internal/cephfs/core"
cerrors "github.com/ceph/ceph-csi/internal/cephfs/errors"
"github.com/ceph/ceph-csi/internal/cephfs/store"
fsutil "github.com/ceph/ceph-csi/internal/cephfs/util"
csicommon "github.com/ceph/ceph-csi/internal/csi-common"
"github.com/ceph/ceph-csi/internal/util"
"github.com/ceph/ceph-csi/internal/util/k8s"
"github.com/ceph/ceph-csi/internal/util/log"
rterrors "github.com/ceph/ceph-csi/internal/util/reftracker/errors"
"github.com/container-storage-interface/spec/lib/go/csi"
"github.com/golang/protobuf/ptypes/timestamp"
"github.com/kubernetes-csi/csi-lib-utils/protosanitizer"
2018-03-05 11:59:47 +00:00
"google.golang.org/grpc/codes"
"google.golang.org/grpc/status"
)
// ControllerServer struct of CEPH CSI driver with supported methods of CSI
// controller server spec.
type ControllerServer struct {
2018-03-05 11:59:47 +00:00
*csicommon.DefaultControllerServer
// A map storing all volumes with ongoing operations so that additional operations
// for that same volume (as defined by VolumeID/volume name) return an Aborted error
VolumeLocks *util.VolumeLocks
// A map storing all snapshots with ongoing operations so that additional operations
// for that same snapshot (as defined by SnapshotID/snapshot name) return an Aborted error
SnapshotLocks *util.VolumeLocks
// A map storing all volumes/snapshots with ongoing operations.
OperationLocks *util.OperationLock
2018-12-19 14:26:16 +00:00
}
// createBackingVolume creates the backing subvolume and on any error cleans up any created entities.
func (cs *ControllerServer) createBackingVolume(
ctx context.Context,
volOptions,
parentVolOpt *store.VolumeOptions,
pvID *store.VolumeIdentifier,
sID *store.SnapshotIdentifier,
) error {
var err error
volClient := core.NewSubVolume(volOptions.GetConnection(), &volOptions.SubVolume, volOptions.ClusterID)
if sID != nil {
return cs.createBackingVolumeFromSnapshotSource(ctx, volOptions, parentVolOpt, volClient, sID)
}
if parentVolOpt != nil {
return cs.createBackingVolumeFromVolumeSource(ctx, parentVolOpt, volClient, pvID)
}
if err = volClient.CreateVolume(ctx); err != nil {
log.ErrorLog(ctx, "failed to create volume %s: %v", volOptions.RequestName, err)
return status.Error(codes.Internal, err.Error())
}
return nil
}
func (cs *ControllerServer) createBackingVolumeFromSnapshotSource(
ctx context.Context,
volOptions *store.VolumeOptions,
parentVolOpt *store.VolumeOptions,
volClient core.SubVolumeClient,
sID *store.SnapshotIdentifier,
) error {
if err := cs.OperationLocks.GetRestoreLock(sID.SnapshotID); err != nil {
log.ErrorLog(ctx, err.Error())
return status.Error(codes.Aborted, err.Error())
}
defer cs.OperationLocks.ReleaseRestoreLock(sID.SnapshotID)
if volOptions.BackingSnapshot {
if err := store.AddSnapshotBackedVolumeRef(ctx, volOptions); err != nil {
log.ErrorLog(ctx, "failed to create snapshot-backed volume from snapshot %s: %v",
sID.FsSnapshotName, err)
return err
}
return nil
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
}
err := volClient.CreateCloneFromSnapshot(ctx, core.Snapshot{
SnapshotID: sID.FsSnapshotName,
SubVolume: &parentVolOpt.SubVolume,
})
if err != nil {
log.ErrorLog(ctx, "failed to create clone from snapshot %s: %v", sID.FsSnapshotName, err)
return err
}
return nil
}
func (cs *ControllerServer) createBackingVolumeFromVolumeSource(
ctx context.Context,
parentVolOpt *store.VolumeOptions,
volClient core.SubVolumeClient,
pvID *store.VolumeIdentifier,
) error {
if err := cs.OperationLocks.GetCloneLock(pvID.VolumeID); err != nil {
log.ErrorLog(ctx, err.Error())
return status.Error(codes.Aborted, err.Error())
}
defer cs.OperationLocks.ReleaseCloneLock(pvID.VolumeID)
if err := volClient.CreateCloneFromSubvolume(ctx, &parentVolOpt.SubVolume); err != nil {
log.ErrorLog(ctx, "failed to create clone from subvolume %s: %v", fsutil.VolumeID(pvID.FsSubvolName), err)
return err
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
}
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
return nil
}
func checkContentSource(
ctx context.Context,
req *csi.CreateVolumeRequest,
cr *util.Credentials,
) (*store.VolumeOptions, *store.VolumeIdentifier, *store.SnapshotIdentifier, error) {
if req.VolumeContentSource == nil {
return nil, nil, nil, nil
}
volumeSource := req.VolumeContentSource
switch volumeSource.Type.(type) {
case *csi.VolumeContentSource_Snapshot:
snapshotID := req.VolumeContentSource.GetSnapshot().GetSnapshotId()
volOpt, _, sid, err := store.NewSnapshotOptionsFromID(ctx, snapshotID, cr)
if err != nil {
if errors.Is(err, cerrors.ErrSnapNotFound) {
return nil, nil, nil, status.Error(codes.NotFound, err.Error())
}
return nil, nil, nil, status.Error(codes.Internal, err.Error())
}
return volOpt, nil, sid, nil
case *csi.VolumeContentSource_Volume:
// Find the volume using the provided VolumeID
volID := req.VolumeContentSource.GetVolume().GetVolumeId()
parentVol, pvID, err := store.NewVolumeOptionsFromVolID(ctx, volID, nil, req.Secrets)
if err != nil {
if !errors.Is(err, cerrors.ErrVolumeNotFound) {
return nil, nil, nil, status.Error(codes.NotFound, err.Error())
}
return nil, nil, nil, status.Error(codes.Internal, err.Error())
}
return parentVol, pvID, nil, nil
}
return nil, nil, nil, status.Errorf(codes.InvalidArgument, "not a proper volume source %v", volumeSource)
}
// checkValidCreateVolumeRequest checks if the request is valid
// CreateVolumeRequest by inspecting the request parameters.
func checkValidCreateVolumeRequest(
vol,
parentVol *store.VolumeOptions,
pvID *store.VolumeIdentifier,
sID *store.SnapshotIdentifier,
req *csi.CreateVolumeRequest,
) error {
switch {
case pvID != nil:
if vol.Size < parentVol.Size {
return fmt.Errorf(
"cannot clone from volume %s: volume size %d is smaller than source volume size %d",
pvID.VolumeID,
parentVol.Size,
vol.Size)
}
if vol.BackingSnapshot {
return errors.New("cloning snapshot-backed volumes is currently not supported")
}
case sID != nil:
if vol.BackingSnapshot {
volCaps := req.GetVolumeCapabilities()
for _, volCap := range volCaps {
mode := volCap.AccessMode.Mode
if mode != csi.VolumeCapability_AccessMode_MULTI_NODE_READER_ONLY &&
mode != csi.VolumeCapability_AccessMode_SINGLE_NODE_READER_ONLY {
return errors.New("backingSnapshot may be used only with read-only access modes")
}
}
}
}
return nil
}
// CreateVolume creates a reservation and the volume in backend, if it is not already present.
// nolint:gocognit,gocyclo,nestif,cyclop // TODO: reduce complexity
func (cs *ControllerServer) CreateVolume(
ctx context.Context,
req *csi.CreateVolumeRequest,
) (*csi.CreateVolumeResponse, error) {
if err := cs.validateCreateVolumeRequest(req); err != nil {
log.ErrorLog(ctx, "CreateVolumeRequest validation failed: %v", err)
2018-03-05 11:59:47 +00:00
return nil, err
}
2019-02-13 12:57:16 +00:00
// Configuration
secret := req.GetSecrets()
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
requestName := req.GetName()
cr, err := util.NewAdminCredentials(secret)
if err != nil {
log.ErrorLog(ctx, "failed to retrieve admin credentials: %v", err)
return nil, status.Error(codes.InvalidArgument, err.Error())
}
defer cr.DeleteCredentials()
// Existence and conflict checks
if acquired := cs.VolumeLocks.TryAcquire(requestName); !acquired {
log.ErrorLog(ctx, util.VolumeOperationAlreadyExistsFmt, requestName)
return nil, status.Errorf(codes.Aborted, util.VolumeOperationAlreadyExistsFmt, requestName)
}
defer cs.VolumeLocks.Release(requestName)
volOptions, err := store.NewVolumeOptions(ctx, requestName, req, cr)
2018-03-05 11:59:47 +00:00
if err != nil {
log.ErrorLog(ctx, "validation and extraction of volume options failed: %v", err)
return nil, status.Error(codes.InvalidArgument, err.Error())
2018-03-05 11:59:47 +00:00
}
defer volOptions.Destroy()
2018-03-05 11:59:47 +00:00
if req.GetCapacityRange() != nil {
volOptions.Size = util.RoundOffCephFSVolSize(req.GetCapacityRange().GetRequiredBytes())
}
parentVol, pvID, sID, err := checkContentSource(ctx, req, cr)
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
if err != nil {
return nil, err
}
if parentVol != nil {
defer parentVol.Destroy()
}
err = checkValidCreateVolumeRequest(volOptions, parentVol, pvID, sID, req)
if err != nil {
return nil, status.Error(codes.InvalidArgument, err.Error())
}
vID, err := store.CheckVolExists(ctx, volOptions, parentVol, pvID, sID, cr)
if err != nil {
if cerrors.IsCloneRetryError(err) {
return nil, status.Error(codes.Aborted, err.Error())
}
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
return nil, status.Error(codes.Internal, err.Error())
}
// TODO return error message if requested vol size greater than found volume return error
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
if vID != nil {
if sID != nil || pvID != nil && !volOptions.BackingSnapshot {
volClient := core.NewSubVolume(volOptions.GetConnection(), &volOptions.SubVolume, volOptions.ClusterID)
err = volClient.ExpandVolume(ctx, volOptions.Size)
if err != nil {
purgeErr := volClient.PurgeVolume(ctx, false)
if purgeErr != nil {
log.ErrorLog(ctx, "failed to delete volume %s: %v", requestName, purgeErr)
// All errors other than ErrVolumeNotFound should return an error back to the caller
if !errors.Is(purgeErr, cerrors.ErrVolumeNotFound) {
return nil, status.Error(codes.Internal, purgeErr.Error())
}
}
errUndo := store.UndoVolReservation(ctx, volOptions, *vID, secret)
if errUndo != nil {
log.WarningLog(ctx, "failed undoing reservation of volume: %s (%s)",
requestName, errUndo)
}
log.ErrorLog(ctx, "failed to expand volume %s: %v", fsutil.VolumeID(vID.FsSubvolName), err)
return nil, status.Error(codes.Internal, err.Error())
}
}
// remove kubernetes csi prefixed parameters.
volumeContext := k8s.RemoveCSIPrefixedParameters(req.GetParameters())
volumeContext["subvolumeName"] = vID.FsSubvolName
volumeContext["subvolumePath"] = volOptions.RootPath
volume := &csi.Volume{
VolumeId: vID.VolumeID,
CapacityBytes: volOptions.Size,
ContentSource: req.GetVolumeContentSource(),
VolumeContext: volumeContext,
}
if volOptions.Topology != nil {
volume.AccessibleTopology = []*csi.Topology{
{
Segments: volOptions.Topology,
},
}
}
return &csi.CreateVolumeResponse{Volume: volume}, nil
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
}
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
// Reservation
vID, err = store.ReserveVol(ctx, volOptions, secret)
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
if err != nil {
return nil, status.Error(codes.Internal, err.Error())
}
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
defer func() {
if err != nil {
if !cerrors.IsCloneRetryError(err) {
errDefer := store.UndoVolReservation(ctx, volOptions, *vID, secret)
if errDefer != nil {
log.WarningLog(ctx, "failed undoing reservation of volume: %s (%s)",
requestName, errDefer)
}
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
}
}
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
}()
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
// Create a volume
err = cs.createBackingVolume(ctx, volOptions, parentVol, pvID, sID)
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
if err != nil {
if cerrors.IsCloneRetryError(err) {
return nil, status.Error(codes.Aborted, err.Error())
}
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
return nil, err
}
if !volOptions.BackingSnapshot {
// Get root path for the created subvolume.
// Note that root path for snapshot-backed volumes has been already set when
// building VolumeOptions.
volClient := core.NewSubVolume(volOptions.GetConnection(), &volOptions.SubVolume, volOptions.ClusterID)
volOptions.RootPath, err = volClient.GetVolumeRootPathCeph(ctx)
if err != nil {
purgeErr := volClient.PurgeVolume(ctx, true)
if purgeErr != nil {
log.ErrorLog(ctx, "failed to delete volume %s: %v", vID.FsSubvolName, purgeErr)
// All errors other than ErrVolumeNotFound should return an error back to the caller
if !errors.Is(purgeErr, cerrors.ErrVolumeNotFound) {
// If the subvolume deletion is failed, we should not cleanup
// the OMAP entry it will stale subvolume in cluster.
// set err=nil so that when we get the request again we can get
// the subvolume info.
err = nil
return nil, status.Error(codes.Internal, purgeErr.Error())
}
}
log.ErrorLog(ctx, "failed to get subvolume path %s: %v", vID.FsSubvolName, err)
return nil, status.Error(codes.Internal, err.Error())
}
}
log.DebugLog(ctx, "cephfs: successfully created backing volume named %s for request name %s",
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
vID.FsSubvolName, requestName)
// remove kubernetes csi prefixed parameters.
volumeContext := k8s.RemoveCSIPrefixedParameters(req.GetParameters())
volumeContext["subvolumeName"] = vID.FsSubvolName
volumeContext["subvolumePath"] = volOptions.RootPath
volume := &csi.Volume{
VolumeId: vID.VolumeID,
CapacityBytes: volOptions.Size,
ContentSource: req.GetVolumeContentSource(),
VolumeContext: volumeContext,
}
if volOptions.Topology != nil {
volume.AccessibleTopology = []*csi.Topology{
{
Segments: volOptions.Topology,
},
}
}
return &csi.CreateVolumeResponse{Volume: volume}, nil
2018-03-05 11:59:47 +00:00
}
// DeleteVolume deletes the volume in backend and its reservation.
func (cs *ControllerServer) DeleteVolume(
ctx context.Context,
req *csi.DeleteVolumeRequest,
) (*csi.DeleteVolumeResponse, error) {
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
if err := cs.validateDeleteVolumeRequest(); err != nil {
log.ErrorLog(ctx, "DeleteVolumeRequest validation failed: %v", err)
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
return nil, err
}
volID := fsutil.VolumeID(req.GetVolumeId())
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
secrets := req.GetSecrets()
// lock out parallel delete operations
if acquired := cs.VolumeLocks.TryAcquire(string(volID)); !acquired {
log.ErrorLog(ctx, util.VolumeOperationAlreadyExistsFmt, volID)
return nil, status.Errorf(codes.Aborted, util.VolumeOperationAlreadyExistsFmt, string(volID))
}
defer cs.VolumeLocks.Release(string(volID))
// lock out volumeID for clone and expand operation
if err := cs.OperationLocks.GetDeleteLock(req.GetVolumeId()); err != nil {
log.ErrorLog(ctx, err.Error())
return nil, status.Error(codes.Aborted, err.Error())
}
defer cs.OperationLocks.ReleaseDeleteLock(req.GetVolumeId())
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
// Find the volume using the provided VolumeID
volOptions, vID, err := store.NewVolumeOptionsFromVolID(ctx, string(volID), nil, secrets)
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
if err != nil {
// if error is ErrPoolNotFound, the pool is already deleted we dont
// need to worry about deleting subvolume or omap data, return success
if errors.Is(err, util.ErrPoolNotFound) {
log.WarningLog(ctx, "failed to get backend volume for %s: %v", string(volID), err)
return &csi.DeleteVolumeResponse{}, nil
}
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
// if error is ErrKeyNotFound, then a previous attempt at deletion was complete
// or partially complete (subvolume and imageOMap are garbage collected already), hence
// return success as deletion is complete
if errors.Is(err, util.ErrKeyNotFound) {
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
return &csi.DeleteVolumeResponse{}, nil
}
log.ErrorLog(ctx, "Error returned from newVolumeOptionsFromVolID: %v", err)
// All errors other than ErrVolumeNotFound should return an error back to the caller
if !errors.Is(err, cerrors.ErrVolumeNotFound) {
return nil, status.Error(codes.Internal, err.Error())
}
// If error is ErrImageNotFound then we failed to find the subvolume, but found the imageOMap
// to lead us to the image, hence the imageOMap needs to be garbage collected, by calling
// unreserve for the same
if acquired := cs.VolumeLocks.TryAcquire(volOptions.RequestName); !acquired {
return nil, status.Errorf(codes.Aborted, util.VolumeOperationAlreadyExistsFmt, volOptions.RequestName)
}
defer cs.VolumeLocks.Release(volOptions.RequestName)
if err = store.UndoVolReservation(ctx, volOptions, *vID, secrets); err != nil {
return nil, status.Error(codes.Internal, err.Error())
}
return &csi.DeleteVolumeResponse{}, nil
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
}
defer volOptions.Destroy()
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
// lock out parallel delete and create requests against the same volume name as we
// cleanup the subvolume and associated omaps for the same
if acquired := cs.VolumeLocks.TryAcquire(volOptions.RequestName); !acquired {
return nil, status.Errorf(codes.Aborted, util.VolumeOperationAlreadyExistsFmt, volOptions.RequestName)
}
defer cs.VolumeLocks.Release(volOptions.RequestName)
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
// Deleting a volume requires admin credentials
cr, err := util.NewAdminCredentials(secrets)
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
if err != nil {
log.ErrorLog(ctx, "failed to retrieve admin credentials: %v", err)
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
return nil, status.Error(codes.InvalidArgument, err.Error())
}
defer cr.DeleteCredentials()
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
if err := cleanUpBackingVolume(ctx, volOptions, vID, cr); err != nil {
return nil, err
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
}
if err := store.UndoVolReservation(ctx, volOptions, *vID, secrets); err != nil {
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
return nil, status.Error(codes.Internal, err.Error())
}
log.DebugLog(ctx, "cephfs: successfully deleted volume %s", volID)
Make CephFS plugin stateless reusing RADOS based journal scheme This is a part of the stateless set of commits for CephCSI. This commit removes the dependency on config maps to store cephFS provisioned volumes, and instead relies on RADOS based objects and keys, and required CSI VolumeID encoding to detect the provisioned volumes. Changes: - Provide backward compatibility to provisioned volumes by older plugin versions (1.0.0 or older) - Remove Create/Delete support for statically provisioned volumes (fixes #382) - Added namespace support to RADOS OMaps and used the same to store RADOS CSI objects and keys in the CephFS metadata pool - Added support to mention fsname for CephFS provisioning (fixes #359) - Changed field name in CSI Identifier to 'location', to denote a pool or fscid - Updated mounter cache to use new scheme - Required Helm manifests are updated - Required documentation and other manifests are updated - Made driver option 'metadatastorage' as optional, as fresh installs do not need to specify the same Testing done: - Create/Mount/Delete PVC - Create/Delete 5 PVCs - Mount version 1.0.0 PVC - Delete version 1.0.0 PV - Mount Statically defined PV/PVC/Pod - Mount Statically defined version 1.0.0 PV/PVC/Pod - Delete Statically defined version 1.0.0 PV/PVC/Pod - Node restart when mounted to test mountcache - Use InstanceID other than 'default' - RBD basic round of tests, as namespace is added to OMaps - csitest against ceph-fs plugin - NOTE: CephFS plugin still does not detect and address already created volumes but of a different size - Test not providing any value to the metadata storage parameter Signed-off-by: ShyamsundarR <srangana@redhat.com>
2019-05-28 19:03:18 +00:00
return &csi.DeleteVolumeResponse{}, nil
}
func cleanUpBackingVolume(
ctx context.Context,
volOptions *store.VolumeOptions,
volID *store.VolumeIdentifier,
cr *util.Credentials,
) error {
if !volOptions.BackingSnapshot {
// Regular volumes need to be purged.
volClient := core.NewSubVolume(volOptions.GetConnection(), &volOptions.SubVolume, volOptions.ClusterID)
if err := volClient.PurgeVolume(ctx, false); err != nil {
log.ErrorLog(ctx, "failed to delete volume %s: %v", volID, err)
if errors.Is(err, cerrors.ErrVolumeHasSnapshots) {
return status.Error(codes.FailedPrecondition, err.Error())
}
if !errors.Is(err, cerrors.ErrVolumeNotFound) {
return status.Error(codes.Internal, err.Error())
}
}
return nil
}
// Snapshot-backed volumes need to un-reference the backing snapshot, and
// the snapshot itself may need to be deleted if its reftracker doesn't
// hold any references anymore.
backingSnapNeedsDelete, err := store.UnrefSnapshotBackedVolume(ctx, volOptions)
if err != nil {
if errors.Is(err, rterrors.ErrObjectOutOfDate) {
return status.Error(codes.Aborted, err.Error())
}
return status.Error(codes.Internal, err.Error())
}
if !backingSnapNeedsDelete {
return nil
}
snapParentVolOptions, _, snapID, err := store.NewSnapshotOptionsFromID(ctx, volOptions.BackingSnapshotID, cr)
if err != nil {
absorbErrs := []error{
util.ErrPoolNotFound,
util.ErrKeyNotFound,
cerrors.ErrSnapNotFound,
cerrors.ErrVolumeNotFound,
}
fatalErr := true
for i := range absorbErrs {
if errors.Is(err, absorbErrs[i]) {
fatalErr = false
break
}
}
if fatalErr {
return status.Error(codes.Internal, err.Error())
}
} else {
snapClient := core.NewSnapshot(
snapParentVolOptions.GetConnection(),
snapID.FsSnapshotName,
&snapParentVolOptions.SubVolume,
)
err = deleteSnapshotAndUndoReservation(ctx, snapClient, snapParentVolOptions, snapID, cr)
if err != nil {
return status.Error(codes.Internal, err.Error())
}
}
return nil
}
// ValidateVolumeCapabilities checks whether the volume capabilities requested
// are supported.
func (cs *ControllerServer) ValidateVolumeCapabilities(
ctx context.Context,
req *csi.ValidateVolumeCapabilitiesRequest,
) (*csi.ValidateVolumeCapabilitiesResponse, error) {
// Cephfs doesn't support Block volume
for _, capability := range req.VolumeCapabilities {
if capability.GetBlock() != nil {
return &csi.ValidateVolumeCapabilitiesResponse{Message: ""}, nil
}
}
return &csi.ValidateVolumeCapabilitiesResponse{
Confirmed: &csi.ValidateVolumeCapabilitiesResponse_Confirmed{
VolumeCapabilities: req.VolumeCapabilities,
},
}, nil
2018-03-05 11:59:47 +00:00
}
// ControllerExpandVolume expands CephFS Volumes on demand based on resizer request.
func (cs *ControllerServer) ControllerExpandVolume(
ctx context.Context,
req *csi.ControllerExpandVolumeRequest,
) (*csi.ControllerExpandVolumeResponse, error) {
if err := cs.validateExpandVolumeRequest(req); err != nil {
log.ErrorLog(ctx, "ControllerExpandVolumeRequest validation failed: %v", err)
return nil, err
}
volID := req.GetVolumeId()
secret := req.GetSecrets()
// lock out parallel delete operations
if acquired := cs.VolumeLocks.TryAcquire(volID); !acquired {
log.ErrorLog(ctx, util.VolumeOperationAlreadyExistsFmt, volID)
return nil, status.Errorf(codes.Aborted, util.VolumeOperationAlreadyExistsFmt, volID)
}
defer cs.VolumeLocks.Release(volID)
// lock out volumeID for clone and delete operation
if err := cs.OperationLocks.GetExpandLock(volID); err != nil {
log.ErrorLog(ctx, err.Error())
return nil, status.Error(codes.Aborted, err.Error())
}
defer cs.OperationLocks.ReleaseExpandLock(volID)
cr, err := util.NewAdminCredentials(secret)
if err != nil {
return nil, status.Error(codes.InvalidArgument, err.Error())
}
defer cr.DeleteCredentials()
volOptions, volIdentifier, err := store.NewVolumeOptionsFromVolID(ctx, volID, nil, secret)
if err != nil {
log.ErrorLog(ctx, "validation and extraction of volume options failed: %v", err)
return nil, status.Error(codes.InvalidArgument, err.Error())
}
defer volOptions.Destroy()
if volOptions.BackingSnapshot {
return nil, status.Error(codes.InvalidArgument, "cannot expand snapshot-backed volume")
}
RoundOffSize := util.RoundOffCephFSVolSize(req.GetCapacityRange().GetRequiredBytes())
volClient := core.NewSubVolume(volOptions.GetConnection(), &volOptions.SubVolume, volOptions.ClusterID)
if err = volClient.ResizeVolume(ctx, RoundOffSize); err != nil {
log.ErrorLog(ctx, "failed to expand volume %s: %v", fsutil.VolumeID(volIdentifier.FsSubvolName), err)
return nil, status.Error(codes.Internal, err.Error())
}
return &csi.ControllerExpandVolumeResponse{
CapacityBytes: RoundOffSize,
NodeExpansionRequired: false,
}, nil
}
// CreateSnapshot creates the snapshot in backend and stores metadata
// in store
// nolint:gocyclo,cyclop // golangci-lint did not catch this earlier, needs to get fixed late
func (cs *ControllerServer) CreateSnapshot(
ctx context.Context,
req *csi.CreateSnapshotRequest,
) (*csi.CreateSnapshotResponse, error) {
if err := cs.validateSnapshotReq(ctx, req); err != nil {
return nil, err
}
cr, err := util.NewAdminCredentials(req.GetSecrets())
if err != nil {
return nil, err
}
defer cr.DeleteCredentials()
clusterData, err := store.GetClusterInformation(req.GetParameters())
if err != nil {
return nil, status.Error(codes.Internal, err.Error())
}
requestName := req.GetName()
sourceVolID := req.GetSourceVolumeId()
// Existence and conflict checks
if acquired := cs.SnapshotLocks.TryAcquire(requestName); !acquired {
log.ErrorLog(ctx, util.SnapshotOperationAlreadyExistsFmt, requestName)
return nil, status.Errorf(codes.Aborted, util.SnapshotOperationAlreadyExistsFmt, requestName)
}
defer cs.SnapshotLocks.Release(requestName)
if err = cs.OperationLocks.GetSnapshotCreateLock(sourceVolID); err != nil {
log.ErrorLog(ctx, err.Error())
return nil, status.Error(codes.Aborted, err.Error())
}
defer cs.OperationLocks.ReleaseSnapshotCreateLock(sourceVolID)
// Find the volume using the provided VolumeID
parentVolOptions, vid, err := store.NewVolumeOptionsFromVolID(ctx, sourceVolID, nil, req.GetSecrets())
if err != nil {
if errors.Is(err, util.ErrPoolNotFound) {
log.WarningLog(ctx, "failed to get backend volume for %s: %v", sourceVolID, err)
return nil, status.Error(codes.NotFound, err.Error())
}
if errors.Is(err, cerrors.ErrVolumeNotFound) {
return nil, status.Error(codes.NotFound, err.Error())
}
return nil, status.Error(codes.Internal, err.Error())
}
defer parentVolOptions.Destroy()
if clusterData.ClusterID != parentVolOptions.ClusterID {
return nil, status.Errorf(
codes.InvalidArgument,
"requested cluster id %s not matching subvolume cluster id %s",
clusterData.ClusterID,
parentVolOptions.ClusterID)
}
if parentVolOptions.BackingSnapshot {
return nil, status.Error(codes.InvalidArgument, "cannot snapshot a snapshot-backed volume")
}
cephfsSnap, genSnapErr := store.GenSnapFromOptions(ctx, req)
if genSnapErr != nil {
return nil, status.Error(codes.Internal, genSnapErr.Error())
}
// lock out parallel snapshot create operations
if acquired := cs.VolumeLocks.TryAcquire(sourceVolID); !acquired {
log.ErrorLog(ctx, util.VolumeOperationAlreadyExistsFmt, sourceVolID)
return nil, status.Errorf(codes.Aborted, util.VolumeOperationAlreadyExistsFmt, sourceVolID)
}
defer cs.VolumeLocks.Release(sourceVolID)
snapName := req.GetName()
sid, snapInfo, err := store.CheckSnapExists(ctx, parentVolOptions, cephfsSnap, cr)
if err != nil {
return nil, status.Error(codes.Internal, err.Error())
}
// check are we able to retrieve the size of parent
// ceph fs subvolume info command got added in 14.2.10 and 15.+
// as we are not able to retrieve the parent size we are rejecting the
// request to create snapshot.
// TODO: For this purpose we could make use of cached clusterAdditionalInfo
// too.
volClient := core.NewSubVolume(parentVolOptions.GetConnection(),
&parentVolOptions.SubVolume,
parentVolOptions.ClusterID)
info, err := volClient.GetSubVolumeInfo(ctx)
if err != nil {
// Check error code value against ErrInvalidCommand to understand the cluster
// support it or not, It's safe to evaluate as the filtering
// is already done from GetSubVolumeInfo() and send out the error here.
if errors.Is(err, cerrors.ErrInvalidCommand) {
return nil, status.Error(
codes.FailedPrecondition,
"subvolume info command not supported in current ceph cluster")
}
if sid != nil {
errDefer := store.UndoSnapReservation(ctx, parentVolOptions, *sid, snapName, cr)
if errDefer != nil {
log.WarningLog(ctx, "failed undoing reservation of snapshot: %s (%s)",
requestName, errDefer)
}
}
return nil, status.Error(codes.Internal, err.Error())
}
if sid != nil {
// check snapshot is protected
protected := true
if !(snapInfo.Protected == core.SnapshotIsProtected) {
snapClient := core.NewSnapshot(parentVolOptions.GetConnection(), sid.FsSnapshotName, &parentVolOptions.SubVolume)
err = snapClient.ProtectSnapshot(ctx)
if err != nil {
protected = false
log.WarningLog(ctx, "failed to protect snapshot of snapshot: %s (%s)",
sid.FsSnapshotName, err)
}
}
return &csi.CreateSnapshotResponse{
Snapshot: &csi.Snapshot{
SizeBytes: info.BytesQuota,
SnapshotId: sid.SnapshotID,
SourceVolumeId: req.GetSourceVolumeId(),
CreationTime: sid.CreationTime,
ReadyToUse: protected,
},
}, nil
}
// Reservation
sID, err := store.ReserveSnap(ctx, parentVolOptions, vid.FsSubvolName, cephfsSnap, cr)
if err != nil {
return nil, status.Error(codes.Internal, err.Error())
}
defer func() {
if err != nil {
errDefer := store.UndoSnapReservation(ctx, parentVolOptions, *sID, snapName, cr)
if errDefer != nil {
log.WarningLog(ctx, "failed undoing reservation of snapshot: %s (%s)",
requestName, errDefer)
}
}
}()
snap, err := doSnapshot(ctx, parentVolOptions, sID.FsSnapshotName)
if err != nil {
return nil, status.Error(codes.Internal, err.Error())
}
return &csi.CreateSnapshotResponse{
Snapshot: &csi.Snapshot{
SizeBytes: info.BytesQuota,
SnapshotId: sID.SnapshotID,
SourceVolumeId: req.GetSourceVolumeId(),
CreationTime: snap.CreationTime,
ReadyToUse: true,
},
}, nil
}
func doSnapshot(
ctx context.Context,
volOpt *store.VolumeOptions,
snapshotName string,
) (core.SnapshotInfo, error) {
snapID := fsutil.VolumeID(snapshotName)
snap := core.SnapshotInfo{}
snapClient := core.NewSnapshot(volOpt.GetConnection(), snapshotName, &volOpt.SubVolume)
err := snapClient.CreateSnapshot(ctx)
if err != nil {
log.ErrorLog(ctx, "failed to create snapshot %s %v", snapID, err)
return snap, err
}
defer func() {
if err != nil {
dErr := snapClient.DeleteSnapshot(ctx)
if dErr != nil {
log.ErrorLog(ctx, "failed to delete snapshot %s %v", snapID, err)
}
}
}()
snap, err = snapClient.GetSnapshotInfo(ctx)
if err != nil {
log.ErrorLog(ctx, "failed to get snapshot info %s %v", snapID, err)
return snap, fmt.Errorf("failed to get snapshot info for snapshot:%s", snapID)
}
var t *timestamp.Timestamp
t, err = fsutil.ParseTime(ctx, snap.CreatedAt)
if err != nil {
return snap, err
}
snap.CreationTime = t
err = snapClient.ProtectSnapshot(ctx)
if err != nil {
log.ErrorLog(ctx, "failed to protect snapshot %s %v", snapID, err)
}
return snap, err
}
func (cs *ControllerServer) validateSnapshotReq(ctx context.Context, req *csi.CreateSnapshotRequest) error {
if err := cs.Driver.ValidateControllerServiceRequest(
csi.ControllerServiceCapability_RPC_CREATE_DELETE_SNAPSHOT); err != nil {
log.ErrorLog(ctx, "invalid create snapshot req: %v", protosanitizer.StripSecrets(req))
return err
}
// Check sanity of request Snapshot Name, Source Volume Id
if req.Name == "" {
return status.Error(codes.NotFound, "snapshot Name cannot be empty")
}
if req.SourceVolumeId == "" {
return status.Error(codes.NotFound, "source Volume ID cannot be empty")
}
return nil
}
// DeleteSnapshot deletes the snapshot in backend and removes the
// snapshot metadata from store.
// nolint:gocyclo,cyclop // TODO: reduce complexity
func (cs *ControllerServer) DeleteSnapshot(
ctx context.Context,
req *csi.DeleteSnapshotRequest,
) (*csi.DeleteSnapshotResponse, error) {
if err := cs.Driver.ValidateControllerServiceRequest(
csi.ControllerServiceCapability_RPC_CREATE_DELETE_SNAPSHOT); err != nil {
log.ErrorLog(ctx, "invalid delete snapshot req: %v", protosanitizer.StripSecrets(req))
return nil, err
}
cr, err := util.NewAdminCredentials(req.GetSecrets())
if err != nil {
return nil, status.Error(codes.InvalidArgument, err.Error())
}
defer cr.DeleteCredentials()
snapshotID := req.GetSnapshotId()
if snapshotID == "" {
return nil, status.Error(codes.InvalidArgument, "snapshot ID cannot be empty")
}
if acquired := cs.SnapshotLocks.TryAcquire(snapshotID); !acquired {
log.ErrorLog(ctx, util.SnapshotOperationAlreadyExistsFmt, snapshotID)
return nil, status.Errorf(codes.Aborted, util.SnapshotOperationAlreadyExistsFmt, snapshotID)
}
defer cs.SnapshotLocks.Release(snapshotID)
// lock out snapshotID for restore operation
if err = cs.OperationLocks.GetDeleteLock(snapshotID); err != nil {
log.ErrorLog(ctx, err.Error())
return nil, status.Error(codes.Aborted, err.Error())
}
defer cs.OperationLocks.ReleaseDeleteLock(snapshotID)
volOpt, snapInfo, sid, err := store.NewSnapshotOptionsFromID(ctx, snapshotID, cr)
if err != nil {
switch {
case errors.Is(err, util.ErrPoolNotFound):
// if error is ErrPoolNotFound, the pool is already deleted we dont
// need to worry about deleting snapshot or omap data, return success
log.WarningLog(ctx, "failed to get backend snapshot for %s: %v", snapshotID, err)
return &csi.DeleteSnapshotResponse{}, nil
case errors.Is(err, util.ErrKeyNotFound):
// if error is ErrKeyNotFound, then a previous attempt at deletion was complete
// or partially complete (snap and snapOMap are garbage collected already), hence return
// success as deletion is complete
return &csi.DeleteSnapshotResponse{}, nil
case errors.Is(err, cerrors.ErrSnapNotFound):
err = store.UndoSnapReservation(ctx, volOpt, *sid, sid.RequestName, cr)
if err != nil {
log.ErrorLog(ctx, "failed to remove reservation for snapname (%s) with backing snap (%s) (%s)",
sid.RequestName, sid.FsSnapshotName, err)
return nil, status.Error(codes.Internal, err.Error())
}
return &csi.DeleteSnapshotResponse{}, nil
case errors.Is(err, cerrors.ErrVolumeNotFound):
// if the error is ErrVolumeNotFound, the subvolume is already deleted
// from backend, Hence undo the omap entries and return success
log.ErrorLog(ctx, "Volume not present")
err = store.UndoSnapReservation(ctx, volOpt, *sid, sid.RequestName, cr)
if err != nil {
log.ErrorLog(ctx, "failed to remove reservation for snapname (%s) with backing snap (%s) (%s)",
sid.RequestName, sid.FsSnapshotName, err)
return nil, status.Error(codes.Internal, err.Error())
}
return &csi.DeleteSnapshotResponse{}, nil
default:
return nil, status.Error(codes.Internal, err.Error())
}
}
defer volOpt.Destroy()
// safeguard against parallel create or delete requests against the same
// name
if acquired := cs.SnapshotLocks.TryAcquire(sid.RequestName); !acquired {
log.ErrorLog(ctx, util.SnapshotOperationAlreadyExistsFmt, sid.RequestName)
return nil, status.Errorf(codes.Aborted, util.VolumeOperationAlreadyExistsFmt, sid.RequestName)
}
defer cs.SnapshotLocks.Release(sid.RequestName)
if snapInfo.HasPendingClones == "yes" {
return nil, status.Errorf(codes.FailedPrecondition, "snapshot %s has pending clones", snapshotID)
}
snapClient := core.NewSnapshot(volOpt.GetConnection(), sid.FsSnapshotName, &volOpt.SubVolume)
if snapInfo.Protected == core.SnapshotIsProtected {
err = snapClient.UnprotectSnapshot(ctx)
if err != nil {
return nil, status.Error(codes.Internal, err.Error())
}
}
needsDelete, err := store.UnrefSelfInSnapshotBackedVolumes(ctx, volOpt, sid.SnapshotID)
if err != nil {
if errors.Is(err, rterrors.ErrObjectOutOfDate) {
return nil, status.Error(codes.Aborted, err.Error())
}
return nil, status.Error(codes.Internal, err.Error())
}
if needsDelete {
err = deleteSnapshotAndUndoReservation(
ctx,
snapClient,
volOpt,
sid,
cr,
)
if err != nil {
return nil, status.Error(codes.Internal, err.Error())
}
}
return &csi.DeleteSnapshotResponse{}, nil
}
func deleteSnapshotAndUndoReservation(
ctx context.Context,
snapClient core.SnapshotClient,
parentVolOptions *store.VolumeOptions,
snapID *store.SnapshotIdentifier,
cr *util.Credentials,
) error {
err := snapClient.DeleteSnapshot(ctx)
if err != nil {
return err
}
err = store.UndoSnapReservation(ctx, parentVolOptions, *snapID, snapID.RequestName, cr)
if err != nil {
log.ErrorLog(ctx, "failed to remove reservation for snapname (%s) with backing snap (%s) (%s)",
snapID.RequestName, snapID.RequestName, err)
return err
}
return nil
}