Files
seaweedfs/weed/shell/command_volume_tier_download.go
T
Chris LuandGitHub 4fb3e22a01 fix(tiering): never delete a shared remote object while replicas still reference it (#9942)
* tiering: stop a shared remote object being deleted while replicas still point at it

A remote-tiered volume's .dat content lives only in one cloud object that all
N replica .vif files point at. Deleting that object while destroying any one
replica, or before a downloaded replica is durable, bricks the survivors.

- volume.tier.move cleanup now deletes old replicas with keepRemoteData=true so
  surviving replicas keep the shared object. Document why the alreadyPlaced
  anchor needs no replica sync (same-object replicas are byte-identical).
- VolumeTierMoveDatFromRemote now fsyncs the downloaded .dat, fsyncs the
  containing directory, trims the .vif (fsynced) and swaps to the local DiskFile
  BEFORE deleting the remote object, on both the keep-remote and delete paths.
  Only the final DeleteFile is gated by keep_remote_dat_file, so a keep-remote
  download leaves the replica served from local disk rather than the shared
  object, and a crash before delete merely leaks the object.
- volume.tier.download keeps the shared object for every replica except the
  last, which deletes it.
- s3 and rclone download paths fsync the .dat before close.

* storage: swap the volume data backend under the data lock

The tier-download swap closed v.DataBackend and assigned the new local DiskFile
without holding dataFileAccessLock, racing concurrent reads/writes (use of a
closed file / nil deref). Add an exported Volume.SwapDataBackend that performs
the close-and-replace under the lock, and call it from the tier download.

* server: skip directory fsync on Windows in the tier download path

os.Open(dir).Sync() is unsupported on Windows and returns an error, which would
fail VolumeTierMoveDatFromRemote entirely there. Skip the directory fsync on
Windows, matching how the storage-side helper tolerates the unsupported case.

* shell: make multi-replica tier.download resilient to already-local replicas

If a multi-replica download is interrupted and retried, a replica made local
in the prior attempt returns "already on local disk", which aborted the whole
command and left the remaining remote replicas dangling. Treat that case as a
skip-and-continue so a retry completes the rest.

* server: assert downloaded .dat content, not just length, in the tier test

A length-only check passes even if the bytes are corrupted; compare the full
content of the local .dat against the original.
2026-06-13 20:09:00 -07:00

199 lines
5.8 KiB
Go

package shell
import (
"context"
"flag"
"fmt"
"io"
"strings"
"github.com/seaweedfs/seaweedfs/weed/pb"
"google.golang.org/grpc"
"github.com/seaweedfs/seaweedfs/weed/operation"
"github.com/seaweedfs/seaweedfs/weed/pb/master_pb"
"github.com/seaweedfs/seaweedfs/weed/pb/volume_server_pb"
"github.com/seaweedfs/seaweedfs/weed/storage/needle"
)
func init() {
Commands = append(Commands, &commandVolumeTierDownload{})
}
type commandVolumeTierDownload struct {
}
func (c *commandVolumeTierDownload) Name() string {
return "volume.tier.download"
}
func (c *commandVolumeTierDownload) Help() string {
return `download the dat file of a volume from a remote tier
volume.tier.download [-collection=""]
volume.tier.download [-collection=""] -volumeId=<volume_id>
The -collection parameter supports regular expressions for pattern matching:
- Use exact match: volume.tier.download -collection="^mybucket$"
- Match multiple buckets: volume.tier.download -collection="bucket.*"
- Match all collections: volume.tier.download -collection=".*"
e.g.:
volume.tier.download -volumeId=7
This command will download the dat file of a volume from a remote tier to a volume server in local cluster.
`
}
func (c *commandVolumeTierDownload) HasTag(CommandTag) bool {
return false
}
func (c *commandVolumeTierDownload) Do(args []string, commandEnv *CommandEnv, writer io.Writer) (err error) {
tierCommand := flag.NewFlagSet(c.Name(), flag.ContinueOnError)
volumeId := tierCommand.Int("volumeId", 0, "the volume id")
collection := tierCommand.String("collection", "", "the collection name")
if err = tierCommand.Parse(args); err != nil {
return nil
}
if err = commandEnv.confirmIsLocked(args); err != nil {
return
}
vid := needle.VolumeId(*volumeId)
// collect topology information
topologyInfo, _, err := collectTopologyInfo(commandEnv, 0)
if err != nil {
return err
}
// volumeId is provided
if vid != 0 {
return doVolumeTierDownload(commandEnv, writer, *collection, vid)
}
// apply to all volumes in the collection
// reusing collectVolumeIdsForEcEncode for now
volumeIds, err := collectRemoteVolumes(topologyInfo, *collection)
if err != nil {
return err
}
fmt.Printf("tier download volumes: %v\n", volumeIds)
for _, vid := range volumeIds {
if err = doVolumeTierDownload(commandEnv, writer, *collection, vid); err != nil {
return err
}
}
return nil
}
func collectRemoteVolumes(topoInfo *master_pb.TopologyInfo, collectionPattern string) (vids []needle.VolumeId, err error) {
// compile regex pattern for collection matching
collectionRegex, err := compileCollectionPattern(collectionPattern)
if err != nil {
return nil, fmt.Errorf("invalid collection pattern '%s': %v", collectionPattern, err)
}
vidMap := make(map[uint32]bool)
eachDataNode(topoInfo, func(dc DataCenterId, rack RackId, dn *master_pb.DataNodeInfo) {
for _, diskInfo := range dn.DiskInfos {
for _, v := range diskInfo.VolumeInfos {
if collectionRegex.MatchString(v.Collection) && v.RemoteStorageKey != "" && v.RemoteStorageName != "" {
vidMap[v.Id] = true
}
}
}
})
for vid := range vidMap {
vids = append(vids, needle.VolumeId(vid))
}
return
}
func doVolumeTierDownload(commandEnv *CommandEnv, writer io.Writer, collection string, vid needle.VolumeId) (err error) {
// find volume location
locations, found := commandEnv.MasterClient.GetLocationsClone(uint32(vid))
if !found {
return fmt.Errorf("volume %d not found", vid)
}
// All replicas point at the same remote object; only the final download may delete
// it. Every earlier replica keeps it so the survivors are not left dangling.
// TODO parallelize this
for i, loc := range locations {
keepRemote := i < len(locations)-1
// copy the .dat file from remote tier to local
err = downloadDatFromRemoteTier(commandEnv.option.GrpcDialOption, writer, needle.VolumeId(vid), collection, loc.ServerAddress(), keepRemote)
if err != nil {
// A replica already made local by a prior interrupted run is not a
// failure; skip it so the remaining remote replicas still download.
if strings.Contains(err.Error(), "already on local disk") {
fmt.Fprintf(writer, "volume %d on %s is already on local disk, skipping\n", vid, loc.Url)
continue
}
return fmt.Errorf("download dat file for volume %d to %s: %v", vid, loc.Url, err)
}
}
return nil
}
func downloadDatFromRemoteTier(grpcDialOption grpc.DialOption, writer io.Writer, volumeId needle.VolumeId, collection string, targetVolumeServer pb.ServerAddress, keepRemote bool) error {
err := operation.WithVolumeServerClient(true, targetVolumeServer, grpcDialOption, func(volumeServerClient volume_server_pb.VolumeServerClient) error {
stream, downloadErr := volumeServerClient.VolumeTierMoveDatFromRemote(context.Background(), &volume_server_pb.VolumeTierMoveDatFromRemoteRequest{
VolumeId: uint32(volumeId),
Collection: collection,
KeepRemoteDatFile: keepRemote,
})
var lastProcessed int64
for {
resp, recvErr := stream.Recv()
if recvErr != nil {
if recvErr == io.EOF {
break
} else {
return recvErr
}
}
processingSpeed := float64(resp.Processed-lastProcessed) / 1024.0 / 1024.0
fmt.Fprintf(writer, "downloaded %.2f%%, %d bytes, %.2fMB/s\n", resp.ProcessedPercentage, resp.Processed, processingSpeed)
lastProcessed = resp.Processed
}
if downloadErr != nil {
return downloadErr
}
_, unmountErr := volumeServerClient.VolumeUnmount(context.Background(), &volume_server_pb.VolumeUnmountRequest{
VolumeId: uint32(volumeId),
})
if unmountErr != nil {
return unmountErr
}
_, mountErr := volumeServerClient.VolumeMount(context.Background(), &volume_server_pb.VolumeMountRequest{
VolumeId: uint32(volumeId),
})
if mountErr != nil {
return mountErr
}
return nil
})
return err
}