Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

rbd: VolumeGroupSnapshot support #4502

Open
wants to merge 14 commits into
base: devel
Choose a base branch
from
Open
Changes from 1 commit
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
214 changes: 214 additions & 0 deletions internal/rbd/group_controller.go
Original file line number Diff line number Diff line change
@@ -0,0 +1,214 @@
/*
Copyright 2024 The Ceph-CSI 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 (
"context"

"github.com/container-storage-interface/spec/lib/go/csi"
"google.golang.org/grpc/codes"
"google.golang.org/grpc/status"

"github.com/ceph/ceph-csi/internal/rbd/types"
"github.com/ceph/ceph-csi/internal/util/log"
)

// CreateVolumeGroupSnapshot receives a list of volume handles and is requested
// to create a list of snapshots that are created at the same time. This is
// similar (although not exactly the same) to consistency groups.
//
// RBD has a limitation that an image can only belong to a single group. It is
// therefore required to create a temporary group, add all images, create the
// group snapshot and remove all images from the group again. This leaves the
// group and its snapshot around, the group snapshot can be inspected to list
// the snapshots of the images.
func (cs *ControllerServer) CreateVolumeGroupSnapshot(
ctx context.Context,
req *csi.CreateVolumeGroupSnapshotRequest,
) (*csi.CreateVolumeGroupSnapshotResponse, error) {
var (
err error
vg types.VolumeGroup
groupSnapshot types.VolumeGroupSnapshot

// the VG and VGS should not have the same name
vgName = req.GetName() + "-vg" // stable temporary name
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What happens if there is already a volumeGroup that was created for VolumeGroup that was used for replication?

vgsName = req.GetName()
)

mgr := NewManager(cs.Driver.GetInstanceID(), req.GetParameters(), req.GetSecrets())
defer mgr.Destroy(ctx)

// resolve all volumes, free them later on
volumes := make([]types.Volume, len(req.GetSourceVolumeIds()))
defer func() {
for _, volume := range volumes {
if vg != nil {
// 'normal' cleanup, remove all images from the group
vgErr := vg.RemoveVolume(ctx, volume)
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This looks good for the group we are creating as part of this function? what about this if the group already created for volumeGroupReplication?

if vgErr != nil {
log.ErrorLog(
ctx,
"failed to remove volume %q from volume group %q: %v",
volume, vg, vgErr)
}
}

// free all allocated volumes
if volume != nil {
volume.Destroy(ctx)
}
}

if vg != nil {
// the VG should always be deleted, volumes can only belong to a single VG
log.DebugLog(ctx, "removing temporary volume group %q", vg)

vgErr := vg.Delete(ctx)
if vgErr != nil {
log.ErrorLog(ctx, "failed to remove temporary volume group %q: %v", vg, vgErr)
}

vg.Destroy(ctx)
}
}()
for i, id := range req.GetSourceVolumeIds() {
var vol types.Volume
vol, err = mgr.GetVolumeByID(ctx, id)
if err != nil {
return nil, status.Errorf(
codes.InvalidArgument,
"failed to find required volume %q for volume group snapshot %q: %s",
id,
vgsName,
err.Error())
}
volumes[i] = vol
}

log.DebugLog(ctx, "all %d Volumes for VolumeGroup %q have been found", len(volumes), vgsName)

groupSnapshot, err = mgr.GetVolumeGroupSnapshotByName(ctx, vgsName)
if groupSnapshot != nil {
csiVGS, csiErr := groupSnapshot.ToCSI(ctx)
if csiErr != nil {
return nil, status.Error(codes.Aborted, csiErr.Error())
}

return &csi.CreateVolumeGroupSnapshotResponse{
GroupSnapshot: csiVGS,
}, nil
}
if err != nil {
log.DebugLog(ctx, "need to create new volume group snapshot, "+
"failed to get existing one with name %q: %v", err)
}

// create a temporary VolumeGroup with a different name
vg, err = mgr.CreateVolumeGroup(ctx, vgName)
if err != nil {
return nil, status.Errorf(
codes.Internal,
"failed to create volume group %q: %s",
vgName,
err.Error())
}

log.DebugLog(ctx, "VolumeGroup %q has been created: %+v", vgName, vg)

// add images to the group
for _, volume := range volumes {
err = vg.AddVolume(ctx, volume)
if err != nil {
return nil, status.Error(codes.Aborted, err.Error())
}
}

groupSnapshot, err = mgr.CreateVolumeGroupSnapshot(ctx, vg, vgsName)
if err != nil {
return nil, status.Error(codes.Aborted, err.Error())
}
defer groupSnapshot.Destroy(ctx)

csiVGS, err := groupSnapshot.ToCSI(ctx)
if err != nil {
return nil, status.Error(codes.Aborted, err.Error())
}

return &csi.CreateVolumeGroupSnapshotResponse{
GroupSnapshot: csiVGS,
}, nil
}

func (cs *ControllerServer) DeleteVolumeGroupSnapshot(
ctx context.Context,
req *csi.DeleteVolumeGroupSnapshotRequest,
) (*csi.DeleteVolumeGroupSnapshotResponse, error) {
// FIXME: more checking of the request in needed
// 1. verify that all snapshots in the request are all snapshots in the group
// 2. delete the group
Comment on lines +161 to +163
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is it going to be done in a followup PR?


mgr := NewManager(cs.Driver.GetInstanceID(), nil, req.GetSecrets())
defer mgr.Destroy(ctx)

groupSnapshot, err := mgr.GetVolumeGroupSnapshotByID(ctx, req.GetGroupSnapshotId())
if err != nil {
return nil, status.Errorf(
codes.Internal,
"failed to get volume group snapshot with id %q: %v",
req.GetGroupSnapshotId(), err)
}
defer groupSnapshot.Destroy(ctx)

err = groupSnapshot.Delete(ctx)
if err != nil {
return nil, status.Errorf(
codes.Internal,
"failed to delete volume group snapshot %q: %v",
groupSnapshot, err)
}
Comment on lines +177 to +183
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is going to take care of deleting all the clones isn't it? just checking to make sure we are good here


return &csi.DeleteVolumeGroupSnapshotResponse{}, nil
}

// GetVolumeGroupSnapshot is sortof optional, only used for
// static/pre-provisioned VolumeGroupSnapshots.
func (cs *ControllerServer) GetVolumeGroupSnapshot(
ctx context.Context,
req *csi.GetVolumeGroupSnapshotRequest,
) (*csi.GetVolumeGroupSnapshotResponse, error) {
mgr := NewManager(cs.Driver.GetInstanceID(), nil, req.GetSecrets())
defer mgr.Destroy(ctx)

groupSnapshot, err := mgr.GetVolumeGroupSnapshotByID(ctx, req.GetGroupSnapshotId())
if err != nil {
return nil, status.Errorf(
codes.Internal,
"failed to get volume group snapshot with id %q: %v",
req.GetGroupSnapshotId(), err)
}
defer groupSnapshot.Destroy(ctx)

csiVGS, err := groupSnapshot.ToCSI(ctx)
if err != nil {
return nil, status.Error(codes.Aborted, err.Error())
}

return &csi.GetVolumeGroupSnapshotResponse{
GroupSnapshot: csiVGS,
}, nil
}