ceph-csi/pkg/rbd/controllerserver.go
Patrick Ohly 25e3a961c3 rbdplugin: idempotent DeleteVolume
When the initial DeleteVolume times out (as it does on slow clusters
due to the low 10 second limit), the external-provisioner calls it
again. The CSI standard requires the second call to succeed if the
volume has been deleted in the meantime. This didn't work because
DeleteVolume returned an error when failing to find the volume info
file:

  rbdplugin: E1008 08:05:35.631783       1 utils.go:100] GRPC error: rbd: open err /var/lib/kubelet/plugins/csi-rbdplugin/controller/csi-rbd-622a252c-cad0-11e8-9112-deadbeef0101.json/open /var/lib/kubelet/plugins/csi-rbdplugin/controller/csi-rbd-622a252c-cad0-11e8-9112-deadbeef0101.json: no such file or directory

The fix is to treat a missing volume info file as "volume already
deleted" and return success. To detect this, the original os error
must be wrapped, otherwise the caller of loadVolInfo cannot determine
the root cause.

Note that further work may be needed to make the driver really
resilient, for example there are probably concurrency issues.
But for now this fixes: #82
2018-10-09 12:08:56 +02:00

422 lines
15 KiB
Go

/*
Copyright 2018 The Kubernetes Authors.
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 rbd
import (
"fmt"
"os"
"os/exec"
"path"
"syscall"
"time"
"github.com/container-storage-interface/spec/lib/go/csi/v0"
"github.com/golang/glog"
"github.com/kubernetes-csi/drivers/pkg/csi-common"
"github.com/pborman/uuid"
"github.com/pkg/errors"
"golang.org/x/net/context"
"google.golang.org/grpc/codes"
"google.golang.org/grpc/status"
)
const (
oneGB = 1073741824
)
type controllerServer struct {
*csicommon.DefaultControllerServer
}
func (cs *controllerServer) CreateVolume(ctx context.Context, req *csi.CreateVolumeRequest) (*csi.CreateVolumeResponse, error) {
if err := cs.Driver.ValidateControllerServiceRequest(csi.ControllerServiceCapability_RPC_CREATE_DELETE_VOLUME); err != nil {
glog.V(3).Infof("invalid create volume req: %v", req)
return nil, err
}
// Check sanity of request Name, Volume Capabilities
if len(req.Name) == 0 {
return nil, status.Error(codes.InvalidArgument, "Volume Name cannot be empty")
}
if req.VolumeCapabilities == nil {
return nil, status.Error(codes.InvalidArgument, "Volume Capabilities cannot be empty")
}
// Need to check for already existing volume name, and if found
// check for the requested capacity and already allocated capacity
if exVol, err := getRBDVolumeByName(req.GetName()); err == nil {
// Since err is nil, it means the volume with the same name already exists
// need to check if the size of exisiting volume is the same as in new
// request
if exVol.VolSize >= int64(req.GetCapacityRange().GetRequiredBytes()) {
// exisiting volume is compatible with new request and should be reused.
// TODO (sbezverk) Do I need to make sure that RBD volume still exists?
return &csi.CreateVolumeResponse{
Volume: &csi.Volume{
Id: exVol.VolID,
CapacityBytes: int64(exVol.VolSize),
Attributes: req.GetParameters(),
},
}, nil
}
return nil, status.Error(codes.AlreadyExists, fmt.Sprintf("Volume with the same name: %s but with different size already exist", req.GetName()))
}
// TODO (sbezverk) Last check for not exceeding total storage capacity
rbdVol, err := getRBDVolumeOptions(req.GetParameters())
if err != nil {
return nil, err
}
// Generating Volume Name and Volume ID, as according to CSI spec they MUST be different
volName := req.GetName()
uniqueID := uuid.NewUUID().String()
if len(volName) == 0 {
volName = rbdVol.Pool + "-dynamic-pvc-" + uniqueID
}
rbdVol.VolName = volName
volumeID := "csi-rbd-" + uniqueID
rbdVol.VolID = volumeID
// Volume Size - Default is 1 GiB
volSizeBytes := int64(oneGB)
if req.GetCapacityRange() != nil {
volSizeBytes = int64(req.GetCapacityRange().GetRequiredBytes())
}
rbdVol.VolSize = volSizeBytes
volSizeGB := int(volSizeBytes / 1024 / 1024 / 1024)
// Check if there is already RBD image with requested name
found, _, _ := rbdStatus(rbdVol, rbdVol.UserId, req.GetControllerCreateSecrets())
if !found {
// if VolumeContentSource is not nil, this request is for snapshot
if req.VolumeContentSource != nil {
snapshot := req.VolumeContentSource.GetSnapshot()
if snapshot == nil {
return nil, status.Error(codes.InvalidArgument, "Volume Snapshot cannot be empty")
}
snapshotID := snapshot.GetId()
if len(snapshotID) == 0 {
return nil, status.Error(codes.InvalidArgument, "Volume Snapshot ID cannot be empty")
}
rbdSnap := &rbdSnapshot{}
if err := loadSnapInfo(snapshotID, path.Join(PluginFolder, "controller-snap"), rbdSnap); err != nil {
return nil, err
}
err = restoreSnapshot(rbdVol, rbdSnap, rbdVol.AdminId, req.GetControllerCreateSecrets())
if err != nil {
return nil, err
}
glog.V(4).Infof("create volume %s from snapshot %s", volName, rbdSnap.SnapName)
} else {
if err := createRBDImage(rbdVol, volSizeGB, rbdVol.AdminId, req.GetControllerCreateSecrets()); err != nil {
if err != nil {
glog.Warningf("failed to create volume: %v", err)
return nil, err
}
}
glog.V(4).Infof("create volume %s", volName)
}
}
// Storing volInfo into a persistent file.
if err := persistVolInfo(volumeID, path.Join(PluginFolder, "controller"), rbdVol); err != nil {
glog.Warningf("rbd: failed to store volInfo with error: %v", err)
}
rbdVolumes[volumeID] = rbdVol
return &csi.CreateVolumeResponse{
Volume: &csi.Volume{
Id: volumeID,
CapacityBytes: int64(volSizeBytes),
Attributes: req.GetParameters(),
},
}, nil
}
func (cs *controllerServer) DeleteVolume(ctx context.Context, req *csi.DeleteVolumeRequest) (*csi.DeleteVolumeResponse, error) {
if err := cs.Driver.ValidateControllerServiceRequest(csi.ControllerServiceCapability_RPC_CREATE_DELETE_VOLUME); err != nil {
glog.Warningf("invalid delete volume req: %v", req)
return nil, err
}
// For now the image get unconditionally deleted, but here retention policy can be checked
volumeID := req.GetVolumeId()
rbdVol := &rbdVolume{}
if err := loadVolInfo(volumeID, path.Join(PluginFolder, "controller"), rbdVol); err != nil {
if os.IsNotExist(errors.Cause(err)) {
// Must have been deleted already. This is not an error (idempotency!).
return &csi.DeleteVolumeResponse{}, nil
}
return nil, err
}
volName := rbdVol.VolName
// Deleting rbd image
glog.V(4).Infof("deleting volume %s", volName)
if err := deleteRBDImage(rbdVol, rbdVol.AdminId, req.GetControllerDeleteSecrets()); err != nil {
// TODO: can we detect "already deleted" situations here and proceed?
glog.V(3).Infof("failed to delete rbd image: %s/%s with error: %v", rbdVol.Pool, volName, err)
return nil, err
}
// Removing persistent storage file for the unmapped volume
if err := deleteVolInfo(volumeID, path.Join(PluginFolder, "controller")); err != nil {
// TODO: we can theoretically end up here when two DeleteVolume calls
// get invoked concurrently. Serialize?
return nil, err
}
delete(rbdVolumes, volumeID)
return &csi.DeleteVolumeResponse{}, nil
}
func (cs *controllerServer) ValidateVolumeCapabilities(ctx context.Context, req *csi.ValidateVolumeCapabilitiesRequest) (*csi.ValidateVolumeCapabilitiesResponse, error) {
for _, cap := range req.VolumeCapabilities {
if cap.GetAccessMode().GetMode() != csi.VolumeCapability_AccessMode_SINGLE_NODE_WRITER {
return &csi.ValidateVolumeCapabilitiesResponse{Supported: false, Message: ""}, nil
}
}
return &csi.ValidateVolumeCapabilitiesResponse{Supported: true, Message: ""}, nil
}
func (cs *controllerServer) ControllerUnpublishVolume(ctx context.Context, req *csi.ControllerUnpublishVolumeRequest) (*csi.ControllerUnpublishVolumeResponse, error) {
return &csi.ControllerUnpublishVolumeResponse{}, nil
}
func (cs *controllerServer) ControllerPublishVolume(ctx context.Context, req *csi.ControllerPublishVolumeRequest) (*csi.ControllerPublishVolumeResponse, error) {
return &csi.ControllerPublishVolumeResponse{}, nil
}
func (cs *controllerServer) CreateSnapshot(ctx context.Context, req *csi.CreateSnapshotRequest) (*csi.CreateSnapshotResponse, error) {
if err := cs.Driver.ValidateControllerServiceRequest(csi.ControllerServiceCapability_RPC_CREATE_DELETE_SNAPSHOT); err != nil {
glog.Warningf("invalid create snapshot req: %v", req)
return nil, err
}
// Check sanity of request Snapshot Name, Source Volume Id
if len(req.Name) == 0 {
return nil, status.Error(codes.InvalidArgument, "Snapshot Name cannot be empty")
}
if len(req.SourceVolumeId) == 0 {
return nil, status.Error(codes.InvalidArgument, "Source Volume ID cannot be empty")
}
// Need to check for already existing snapshot name, and if found
// check for the requested source volume id and already allocated source volume id
if exSnap, err := getRBDSnapshotByName(req.GetName()); err == nil {
if req.SourceVolumeId == exSnap.SourceVolumeID {
return &csi.CreateSnapshotResponse{
Snapshot: &csi.Snapshot{
SizeBytes: exSnap.SizeBytes,
Id: exSnap.SnapID,
SourceVolumeId: exSnap.SourceVolumeID,
CreatedAt: exSnap.CreatedAt,
Status: &csi.SnapshotStatus{
Type: csi.SnapshotStatus_READY,
},
},
}, nil
}
return nil, status.Error(codes.AlreadyExists, fmt.Sprintf("Snapshot with the same name: %s but with different source volume id already exist", req.GetName()))
}
rbdSnap, err := getRBDSnapshotOptions(req.GetParameters())
if err != nil {
return nil, err
}
// Generating Snapshot Name and Snapshot ID, as according to CSI spec they MUST be different
snapName := req.GetName()
uniqueID := uuid.NewUUID().String()
rbdVolume, err := getRBDVolumeByID(req.GetSourceVolumeId())
if err != nil {
return nil, status.Error(codes.NotFound, fmt.Sprintf("Source Volume ID %s cannot found", req.GetSourceVolumeId()))
}
if !hasSnapshotFeature(rbdVolume.ImageFeatures) {
return nil, fmt.Errorf("Volume(%s) has not snapshot feature(layering)", req.GetSourceVolumeId())
}
rbdSnap.VolName = rbdVolume.VolName
rbdSnap.SnapName = snapName
snapshotID := "csi-rbd-" + rbdVolume.VolName + "-snap-" + uniqueID
rbdSnap.SnapID = snapshotID
rbdSnap.SourceVolumeID = req.GetSourceVolumeId()
rbdSnap.SizeBytes = rbdVolume.VolSize
err = createSnapshot(rbdSnap, rbdSnap.AdminId, req.GetCreateSnapshotSecrets())
// if we already have the snapshot, return the snapshot
if err != nil {
if exitErr, ok := err.(*exec.ExitError); ok {
if status, ok := exitErr.Sys().(syscall.WaitStatus); ok {
if status.ExitStatus() == int(syscall.EEXIST) {
glog.Warningf("Snapshot with the same name: %s, we return this.", req.GetName())
} else {
glog.Warningf("failed to create snapshot: %v", err)
return nil, err
}
} else {
glog.Warningf("failed to create snapshot: %v", err)
return nil, err
}
} else {
glog.Warningf("failed to create snapshot: %v", err)
return nil, err
}
} else {
glog.V(4).Infof("create snapshot %s", snapName)
err = protectSnapshot(rbdSnap, rbdSnap.AdminId, req.GetCreateSnapshotSecrets())
if err != nil {
err = deleteSnapshot(rbdSnap, rbdSnap.AdminId, req.GetCreateSnapshotSecrets())
if err != nil {
return nil, fmt.Errorf("snapshot is created but failed to protect and delete snapshot: %v", err)
}
return nil, fmt.Errorf("Snapshot is created but failed to protect snapshot")
}
}
rbdSnap.CreatedAt = time.Now().UnixNano()
// Storing snapInfo into a persistent file.
if err := persistSnapInfo(snapshotID, path.Join(PluginFolder, "controller-snap"), rbdSnap); err != nil {
glog.Warningf("rbd: failed to store snapInfo with error: %v", err)
// Unprotect snapshot
err := unprotectSnapshot(rbdSnap, rbdSnap.AdminId, req.GetCreateSnapshotSecrets())
if err != nil {
return nil, status.Error(codes.Unknown, fmt.Sprintf("This Snapshot should be removed but failed to unprotect snapshot: %s/%s with error: %v", rbdSnap.Pool, rbdSnap.SnapName, err))
}
// Deleting snapshot
glog.V(4).Infof("deleting Snaphot %s", rbdSnap.SnapName)
if err := deleteSnapshot(rbdSnap, rbdSnap.AdminId, req.GetCreateSnapshotSecrets()); err != nil {
return nil, status.Error(codes.Unknown, fmt.Sprintf("This Snapshot should be removed but failed to delete snapshot: %s/%s with error: %v", rbdSnap.Pool, rbdSnap.SnapName, err))
}
return nil, err
}
rbdSnapshots[snapshotID] = rbdSnap
return &csi.CreateSnapshotResponse{
Snapshot: &csi.Snapshot{
SizeBytes: rbdSnap.SizeBytes,
Id: snapshotID,
SourceVolumeId: req.GetSourceVolumeId(),
CreatedAt: rbdSnap.CreatedAt,
Status: &csi.SnapshotStatus{
Type: csi.SnapshotStatus_READY,
},
},
}, nil
}
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 {
glog.Warningf("invalid delete snapshot req: %v", req)
return nil, err
}
snapshotID := req.GetSnapshotId()
if len(snapshotID) == 0 {
return nil, status.Error(codes.InvalidArgument, "Snapshot ID cannot be empty")
}
rbdSnap := &rbdSnapshot{}
if err := loadSnapInfo(snapshotID, path.Join(PluginFolder, "controller-snap"), rbdSnap); err != nil {
return nil, err
}
// Unprotect snapshot
err := unprotectSnapshot(rbdSnap, rbdSnap.AdminId, req.GetDeleteSnapshotSecrets())
if err != nil {
return nil, status.Error(codes.FailedPrecondition, fmt.Sprintf("failed to unprotect snapshot: %s/%s with error: %v", rbdSnap.Pool, rbdSnap.SnapName, err))
}
// Deleting snapshot
glog.V(4).Infof("deleting Snaphot %s", rbdSnap.SnapName)
if err := deleteSnapshot(rbdSnap, rbdSnap.AdminId, req.GetDeleteSnapshotSecrets()); err != nil {
return nil, status.Error(codes.FailedPrecondition, fmt.Sprintf("failed to delete snapshot: %s/%s with error: %v", rbdSnap.Pool, rbdSnap.SnapName, err))
}
// Removing persistent storage file for the unmapped snapshot
if err := deleteSnapInfo(snapshotID, path.Join(PluginFolder, "controller-snap")); err != nil {
return nil, err
}
delete(rbdSnapshots, snapshotID)
return &csi.DeleteSnapshotResponse{}, nil
}
func (cs *controllerServer) ListSnapshots(ctx context.Context, req *csi.ListSnapshotsRequest) (*csi.ListSnapshotsResponse, error) {
if err := cs.Driver.ValidateControllerServiceRequest(csi.ControllerServiceCapability_RPC_LIST_SNAPSHOTS); err != nil {
glog.Warningf("invalid list snapshot req: %v", req)
return nil, err
}
sourceVolumeId := req.GetSourceVolumeId()
// TODO (sngchlko) list with token
// list only a specific snapshot which has snapshot ID
if snapshotID := req.GetSnapshotId(); len(snapshotID) != 0 {
if rbdSnap, ok := rbdSnapshots[snapshotID]; ok {
// if source volume ID also set, check source volume id on the cache.
if len(sourceVolumeId) != 0 && rbdSnap.SourceVolumeID != sourceVolumeId {
return nil, status.Error(codes.Unknown, fmt.Sprintf("Requested Source Volume ID %s is different from %s", sourceVolumeId, rbdSnap.SourceVolumeID))
}
return &csi.ListSnapshotsResponse{
Entries: []*csi.ListSnapshotsResponse_Entry{
{
Snapshot: &csi.Snapshot{
SizeBytes: rbdSnap.SizeBytes,
Id: rbdSnap.SnapID,
SourceVolumeId: rbdSnap.SourceVolumeID,
CreatedAt: rbdSnap.CreatedAt,
Status: &csi.SnapshotStatus{
Type: csi.SnapshotStatus_READY,
},
},
},
},
}, nil
} else {
return nil, status.Error(codes.NotFound, fmt.Sprintf("Snapshot ID %s cannot found", snapshotID))
}
}
entries := []*csi.ListSnapshotsResponse_Entry{}
for _, rbdSnap := range rbdSnapshots {
// if source volume ID also set, check source volume id on the cache.
if len(sourceVolumeId) != 0 && rbdSnap.SourceVolumeID != sourceVolumeId {
continue
}
entries = append(entries, &csi.ListSnapshotsResponse_Entry{
Snapshot: &csi.Snapshot{
SizeBytes: rbdSnap.SizeBytes,
Id: rbdSnap.SnapID,
SourceVolumeId: rbdSnap.SourceVolumeID,
CreatedAt: rbdSnap.CreatedAt,
Status: &csi.SnapshotStatus{
Type: csi.SnapshotStatus_READY,
},
},
})
}
return &csi.ListSnapshotsResponse{
Entries: entries,
}, nil
}