seaweedfs/weed/storage/store_ec.go

package storage

import (
	"context"
	"errors"
	"fmt"
	"io"
	"os"
	"slices"
	"sync"
	"time"

	"github.com/klauspost/reedsolomon"

	"github.com/seaweedfs/seaweedfs/weed/glog"
	"github.com/seaweedfs/seaweedfs/weed/operation"
	"github.com/seaweedfs/seaweedfs/weed/pb"
	"github.com/seaweedfs/seaweedfs/weed/pb/master_pb"
	"github.com/seaweedfs/seaweedfs/weed/pb/volume_server_pb"
	"github.com/seaweedfs/seaweedfs/weed/stats"
	"github.com/seaweedfs/seaweedfs/weed/storage/erasure_coding"
	"github.com/seaweedfs/seaweedfs/weed/storage/needle"
	"github.com/seaweedfs/seaweedfs/weed/storage/types"
)

// errShardNotLocal indicates that the requested EC shard is simply not
// stored on this volume server. It is expected during normal reads when
// shards are spread across multiple servers, so callers should not log
// it as an error.
var errShardNotLocal = errors.New("ec shard not on this server")

// FindEcShardTargetLocation returns the disk that should receive a new
// shard / index file for (collection, vid). The selection order is:
//
//  1. a disk that already has the EC volume mounted (in-memory state),
//  2. a disk that owns the .ecx file on disk (volume not mounted yet),
//  3. any HDD with free space,
//  4. any disk with free space.
//
// Step 2 is the missing primitive that pinned subsequent shards to the
// first-shard disk during ec.rebuild. ec.rebuild only sets CopyEcxFile=true
// for the first shard, then relies on auto-select to land later shards on
// the same disk. Without an on-disk check, FindEcVolume returns nothing
// (no mount yet) and the fallback picks "any HDD with free space" — which
// can split shards from their index files across disks of the same node
// and lose them at startup. See issue #9212 and the orphan-shard
// reconciliation in #9244.
//
// dataShardCount is the data-shard count for this volume's EC layout (10
// for the OSS default, but custom ratios are supported via .vif). Callers
// pass it explicitly so this helper stays free of package-level constants
// — easier to mirror into builds that ship a different default ratio.
//
// Implementation walks s.Locations once and scores each disk by tier; the
// highest-tier disk wins, ties broken by free count. The earlier waterfall
// across four FindFreeLocation passes was equivalent but acquired
// volumesLock and ecVolumesLock RLocks (via VolumesLen / EcShardCount) up
// to four times per disk per call.
func (s *Store) FindEcShardTargetLocation(collection string, vid needle.VolumeId, dataShardCount int) *DiskLocation {
	const (
		tierAnyDisk = iota + 1
		tierHDD
		tierEcxOnDisk
		tierMounted
	)

	var (
		best     *DiskLocation
		bestTier int
		bestFree int32
	)
	for _, loc := range s.Locations {
		if loc.isDiskSpaceLow.Load() {
			continue
		}
		freeCount := ecFreeShardCount(loc, dataShardCount)
		if freeCount <= 0 {
			continue
		}
		tier := tierAnyDisk
		if loc.DiskType == types.HardDriveType {
			tier = tierHDD
		}
		if loc.HasEcxFileOnDisk(collection, vid) {
			tier = tierEcxOnDisk
		}
		if _, mounted := loc.FindEcVolume(vid); mounted {
			tier = tierMounted
		}
		if best == nil || tier > bestTier || (tier == bestTier && freeCount > bestFree) {
			best = loc
			bestTier = tier
			bestFree = freeCount
		}
	}
	return best
}

// ecFreeShardCount returns the free EC shard capacity of loc, expressed
// in shard slots (not volume-equivalent slots). dataShardCount is the
// data-shard count of the EC layout being placed — see
// FindEcShardTargetLocation's docstring for why it's a parameter.
//
// FindFreeLocation in store.go does the same math but divides by
// DataShardsCount at the end. That truncation can exclude a disk that
// still has room for several individual shards (e.g. MaxVolumeCount=1,
// EcShardCount=1, dataShardCount=10 → reports 0 despite 9 free shard
// slots), which in this helper would re-route subsequent shards off the
// .ecx-owning disk and re-introduce the orphan-shard layout #9212 is
// trying to prevent. So we keep the result in shard slots throughout.
//
// MaxVolumeCount == 0 is the "unlimited" sentinel used elsewhere in the
// store (see hasFreeDiskLocation). Reporting a synthetic large free
// count keeps unlimited disks eligible while still letting tie-breaks
// prefer the less-loaded one.
func ecFreeShardCount(loc *DiskLocation, dataShardCount int) int32 {
	if dataShardCount <= 0 {
		return 0
	}
	if loc.MaxVolumeCount <= 0 {
		const unlimitedFree = int32(1 << 30)
		used := int32(loc.VolumesLen())*int32(dataShardCount) + int32(loc.EcShardCount())
		if used >= unlimitedFree {
			return 1
		}
		return unlimitedFree - used
	}
	free := (loc.MaxVolumeCount - int32(loc.VolumesLen())) * int32(dataShardCount)
	free -= int32(loc.EcShardCount())
	if free < 0 {
		return 0
	}
	return free
}

func (s *Store) CollectErasureCodingHeartbeat() *master_pb.Heartbeat {
	var ecShardMessages []*master_pb.VolumeEcShardInformationMessage
	collectionEcShardSize := make(map[string]int64)
	for diskId, location := range s.Locations {
		location.ecVolumesLock.RLock()
		for _, ecShards := range location.ecVolumes {
			ecShardMessages = append(ecShardMessages, ecShards.ToVolumeEcShardInformationMessage(uint32(diskId))...)

			for _, ecShard := range ecShards.Shards {
				collectionEcShardSize[ecShards.Collection] += ecShard.Size()
			}
		}
		location.ecVolumesLock.RUnlock()
	}

	for col, size := range collectionEcShardSize {
		stats.VolumeServerDiskSizeGauge.WithLabelValues(col, "ec").Set(float64(size))
	}

	return &master_pb.Heartbeat{
		EcShards:      ecShardMessages,
		HasNoEcShards: len(ecShardMessages) == 0,
	}

}

func (s *Store) MountEcShards(collection string, vid needle.VolumeId, shardId erasure_coding.ShardId, sourceDiskType string) error {
	// The .ecx index file may live on a different disk than the one
	// holding the .ec?? shard being mounted: ec.balance / ec.rebuild can
	// place the .ecx on one local disk while later distributing shards
	// across sibling disks of the same volume server. The per-disk
	// IdxDirectory used by LoadEcShard would ENOENT the .ecx, so look up
	// the .ecx owner across all DiskLocations once and route NewEcVolume
	// at the directory that actually has the file. A 0-byte .ecx is
	// treated as missing here (writeToFile can leave a stub on a failed
	// EC distribute) so we still scan the rest of the disks.
	ecxIdxDir, ecxFound := s.findEcxIdxDirForVolume(collection, vid)

	// Collect failures so an all-disks-fail return reports every disk we
	// tried rather than just the first one. Before this loop reordered
	// itself to keep going after the first non-ENOENT error, a single
	// shard-on-disk-without-.ecx situation would bail the loop and the
	// operator saw "cannot open ec volume index" naming exactly one disk
	// even when others held a valid index.
	type diskError struct {
		dir string
		err error
	}
	var failures []diskError

	for diskId, location := range s.Locations {
		idxDir := location.IdxDirectory
		if ecxFound {
			// Fast path: if findEcxIdxDirForVolume already pointed at
			// one of this disk's directories, the disk owns the .ecx
			// and the local IdxDirectory is the right answer — skip
			// the HasEcxFileOnDisk stat. Only fall back to the sibling
			// disk's idxDir when this disk's directories are neither.
			if location.IdxDirectory != ecxIdxDir && location.Directory != ecxIdxDir {
				if !location.HasEcxFileOnDisk(collection, vid) {
					idxDir = ecxIdxDir
				}
			}
		}
		ecVolume, err := location.loadEcShardWithIdxDir(collection, vid, shardId, idxDir)
		if err == nil {
			glog.V(0).Infof("MountEcShards %d.%d on disk ID %d", vid, shardId, diskId)

			// Apply the orchestrator-supplied source disk type so the EC
			// volume reports under it instead of the location's. Empty means
			// "fall back to location's disk type" (#9423).
			if sourceDiskType != "" {
				ecVolume.SetDiskType(types.ToDiskType(sourceDiskType))
			}

			si := erasure_coding.NewShardsInfo()
			si.Set(erasure_coding.NewShardInfo(shardId, erasure_coding.ShardSize(ecVolume.ShardSize())))
			s.NewEcShardsChan <- &master_pb.VolumeEcShardInformationMessage{
				Id:          uint32(vid),
				Collection:  collection,
				EcIndexBits: uint32(si.Bitmap()),
				ShardSizes:  si.SizesInt64(),
				DiskType:    string(ecVolume.DiskType()),
				ExpireAtSec: ecVolume.ExpireAtSec,
				DiskId:      uint32(diskId),
			}
			return nil
		}
		if errors.Is(err, os.ErrNotExist) {
			// Shard or index not on this disk; another disk may own it.
			continue
		}
		failures = append(failures, diskError{dir: location.Directory, err: err})
	}

	if len(failures) == 0 {
		// No disk had the shard or the index; this volume server is not
		// holding any artefacts for the requested shard. Name what we
		// scanned for so the operator can tell "no .ecx anywhere" apart
		// from "shard not on this server".
		if !ecxFound {
			return fmt.Errorf("MountEcShards %d.%d: no .ecx index found on any local disk", vid, shardId)
		}
		return fmt.Errorf("MountEcShards %d.%d not found on disk", vid, shardId)
	}
	// Some disks returned a real (non-ENOENT) error. Report them all so
	// the caller can see whether the failures cluster around one disk
	// (likely hardware) or are spread out (likely a config problem).
	var b []byte
	for i, f := range failures {
		if i > 0 {
			b = append(b, "; "...)
		}
		b = append(b, fmt.Sprintf("%s: %v", f.dir, f.err)...)
	}
	return fmt.Errorf("MountEcShards %d.%d load failures: %s", vid, shardId, string(b))
}

func (s *Store) UnmountEcShards(vid needle.VolumeId, shardId erasure_coding.ShardId) error {
	// Walk every disk: a split-disk reconciled volume can mount the same vid on
	// more than one disk, so a first-match unmount would leave a sibling copy
	// mounted and heartbeating. Emit one deletion delta per disk.
	unmountedAny := false
	var lastErr error
	for diskId, location := range s.Locations {
		ecShard, found := location.FindEcShard(vid, shardId)
		if !found {
			continue
		}
		if deleted := location.UnloadEcShard(vid, shardId); deleted {
			si := erasure_coding.NewShardsInfo()
			si.Set(erasure_coding.NewShardInfo(shardId, 0))
			s.DeletedEcShardsChan <- &master_pb.VolumeEcShardInformationMessage{
				Id:          uint32(vid),
				Collection:  ecShard.Collection,
				EcIndexBits: si.Bitmap(),
				ShardSizes:  si.SizesInt64(),
				DiskType:    string(ecShard.DiskType),
				DiskId:      uint32(diskId),
			}
			glog.V(0).Infof("UnmountEcShards %d.%d disk_id:%d", vid, shardId, diskId)
			unmountedAny = true
		} else {
			lastErr = fmt.Errorf("UnmountEcShards %d.%d not found on disk %d", vid, shardId, diskId)
		}
	}

	// nil when no disk held the shard (idempotent re-unmount).
	if !unmountedAny {
		return lastErr
	}
	return nil
}

func (s *Store) findEcShard(vid needle.VolumeId, shardId erasure_coding.ShardId) (diskId uint32, shard *erasure_coding.EcVolumeShard, found bool) {
	for diskId, location := range s.Locations {
		if v, found := location.FindEcShard(vid, shardId); found {
			return uint32(diskId), v, found
		}
	}
	return 0, nil, false
}

// FindEcShard returns the shard if any DiskLocation on this server holds it,
// along with that disk's id.
func (s *Store) FindEcShard(vid needle.VolumeId, shardId erasure_coding.ShardId) (diskId uint32, shard *erasure_coding.EcVolumeShard, found bool) {
	return s.findEcShard(vid, shardId)
}

// FindEcVolumeWithShard returns the EcVolume on the disk that owns the given
// shard, plus the shard. The read guard must check the identity of the volume
// that owns the bytes served: on a multi-disk server one vid can hold shards
// from different encode runs across disks, so a first-match volume can differ.
func (s *Store) FindEcVolumeWithShard(vid needle.VolumeId, shardId erasure_coding.ShardId) (*erasure_coding.EcVolume, *erasure_coding.EcVolumeShard, bool) {
	for _, location := range s.Locations {
		if shard, found := location.FindEcShard(vid, shardId); found {
			if ev, ok := location.FindEcVolume(vid); ok {
				return ev, shard, true
			}
		}
	}
	return nil, nil, false
}

func (s *Store) FindEcVolume(vid needle.VolumeId) (*erasure_coding.EcVolume, bool) {
	for _, location := range s.Locations {
		if s, found := location.FindEcVolume(vid); found {
			return s, true
		}
	}
	return nil, false
}

// FindEcVolumeDiskIds returns every disk_id on this store that has an
// EcVolume entry for the given volume. Useful for diagnostic logging
// when a single FindEcVolume hit hides which disk is actually holding
// the mount (e.g., the ReceiveFile mounted-volume guard).
func (s *Store) FindEcVolumeDiskIds(vid needle.VolumeId) []uint32 {
	var ids []uint32
	for diskId, location := range s.Locations {
		if _, found := location.FindEcVolume(vid); found {
			ids = append(ids, uint32(diskId))
		}
	}
	return ids
}

// shardFiles is a list of shard files, which is used to return the shard locations
func (s *Store) CollectEcShards(vid needle.VolumeId, shardFileNames []string) (ecVolume *erasure_coding.EcVolume, found bool) {
	for _, location := range s.Locations {
		if s, foundShards := location.CollectEcShards(vid, shardFileNames); foundShards {
			ecVolume = s
			found = true
		}
	}
	return
}

func (s *Store) DestroyEcVolume(vid needle.VolumeId) {
	for _, location := range s.Locations {
		location.DestroyEcVolume(vid)
	}
}

// UnloadEcVolume drops any in-memory EcVolume for vid from every disk and closes
// its fds without deleting files, so a following unlink frees the inodes.
func (s *Store) UnloadEcVolume(vid needle.VolumeId) {
	for _, location := range s.Locations {
		location.unloadEcVolume(vid)
	}
}

func (s *Store) ReadEcShardNeedle(vid needle.VolumeId, n *needle.Needle, onReadSizeFn func(size types.Size)) (int, error) {
	for _, location := range s.Locations {
		if localEcVolume, found := location.FindEcVolume(vid); found {

			offset, size, intervals, err := localEcVolume.LocateEcShardNeedle(n.Id, localEcVolume.Version)
			if err != nil {
				return 0, fmt.Errorf("locate in local ec volume: %w", err)
			}
			if size.IsDeleted() {
				return 0, ErrorDeleted
			}

			if onReadSizeFn != nil {
				onReadSizeFn(size)
			}

			glog.V(3).Infof("read ec volume %d offset %d size %d intervals:%+v", vid, offset.ToActualOffset(), size, intervals)

			if len(intervals) > 1 {
				glog.V(3).Infof("ReadEcShardNeedle needle id %s intervals:%+v", n.String(), intervals)
			}
			bytes, isDeleted, err := s.readEcShardIntervals(n.Id, localEcVolume, intervals)
			if err != nil {
				return 0, fmt.Errorf("ReadEcShardIntervals: %w", err)
			}
			if isDeleted {
				return 0, ErrorDeleted
			}

			err = n.ReadBytes(bytes, offset.ToActualOffset(), size, localEcVolume.Version)
			if err != nil {
				return 0, fmt.Errorf("ec volume %d needle %s offset %d size %d: %w", vid, n.String(), offset.ToActualOffset(), size, err)
			}

			return len(bytes), nil
		}
	}
	return 0, fmt.Errorf("ec shard %d not found", vid)
}

func (s *Store) IntervalToShardIdAndOffset(iv erasure_coding.Interval) (erasure_coding.ShardId, int64) {
	return iv.ToShardIdAndOffset(erasure_coding.ErasureCodingLargeBlockSize, erasure_coding.ErasureCodingSmallBlockSize)
}

func (s *Store) readEcShardIntervals(needleId types.NeedleId, ecVolume *erasure_coding.EcVolume, intervals []erasure_coding.Interval) (data []byte, is_deleted bool, err error) {
	if err = s.cachedLookupEcShardLocations(ecVolume); err != nil {
		return nil, false, fmt.Errorf("failed to locate shard via master grpc %s: %v", s.MasterAddress, err)
	}

	for i, interval := range intervals {
		if d, isDeleted, e := s.readOneEcShardInterval(needleId, ecVolume, interval); e != nil {
			return nil, isDeleted, e
		} else {
			if isDeleted {
				is_deleted = true
			}
			if i == 0 {
				data = d
			} else {
				data = append(data, d...)
			}
		}
	}
	return
}

func (s *Store) readOneEcShardInterval(needleId types.NeedleId, ecVolume *erasure_coding.EcVolume, interval erasure_coding.Interval) (data []byte, is_deleted bool, err error) {
	shardId, actualOffset := s.IntervalToShardIdAndOffset(interval)
	data = make([]byte, interval.Size)

	// try local read
	err = s.readLocalEcShardInterval(ecVolume, shardId, data, actualOffset)
	if err == nil {
		return
	}
	if errors.Is(err, errShardNotLocal) {
		// expected when shards are spread across servers; fall through to remote read
		glog.V(4).Infof("ec shard %d.%d not local, will try remote", ecVolume.VolumeId, shardId)
	} else {
		glog.V(0).Infof("read local ec shard %d.%d offset %d: %v", ecVolume.VolumeId, shardId, actualOffset, err)
	}

	ecVolume.ShardLocationsLock.RLock()
	sourceDataNodes, hasShardIdLocation := ecVolume.ShardLocations[shardId]
	ecVolume.ShardLocationsLock.RUnlock()

	// try reading directly
	if hasShardIdLocation {
		_, is_deleted, err = s.readRemoteEcShardInterval(sourceDataNodes, needleId, ecVolume.VolumeId, shardId, data, actualOffset, ecVolume.EncodeTsNs)
		if err == nil {
			return
		}
		glog.V(0).Infof("read remote ec shard %d.%d locations: %v", ecVolume.VolumeId, shardId, err)
	}

	// try reading by recovering from other shards
	_, is_deleted, err = s.recoverOneRemoteEcShardInterval(needleId, ecVolume, shardId, data, actualOffset)
	if err == nil {
		return
	}
	glog.V(0).Infof("recover ec shard %d.%d : %v", ecVolume.VolumeId, shardId, err)

	return
}

func forgetShardId(ecVolume *erasure_coding.EcVolume, shardId erasure_coding.ShardId) {
	// failed to access the source data nodes, clear it up
	ecVolume.ShardLocationsLock.Lock()
	delete(ecVolume.ShardLocations, shardId)
	ecVolume.ShardLocationsLock.Unlock()
}

func (s *Store) cachedLookupEcShardLocations(ecVolume *erasure_coding.EcVolume) (err error) {

	shardCount := len(ecVolume.ShardLocations)
	if shardCount < erasure_coding.DataShardsCount &&
		ecVolume.ShardLocationsRefreshTime.Add(11*time.Second).After(time.Now()) ||
		shardCount == erasure_coding.TotalShardsCount &&
			ecVolume.ShardLocationsRefreshTime.Add(37*time.Minute).After(time.Now()) ||
		shardCount >= erasure_coding.DataShardsCount &&
			ecVolume.ShardLocationsRefreshTime.Add(7*time.Minute).After(time.Now()) {
		// still fresh
		return nil
	}

	glog.V(3).Infof("lookup and cache ec volume %d locations", ecVolume.VolumeId)

	err = operation.WithMasterServerClient(context.Background(), false, s.MasterAddress, s.grpcDialOption, func(masterClient master_pb.SeaweedClient) error {
		req := &master_pb.LookupEcVolumeRequest{
			VolumeId: uint32(ecVolume.VolumeId),
		}
		resp, err := masterClient.LookupEcVolume(context.Background(), req)
		if err != nil {
			return fmt.Errorf("lookup ec volume %d: %v", ecVolume.VolumeId, err)
		}
		if len(resp.ShardIdLocations) < erasure_coding.DataShardsCount {
			return fmt.Errorf("only %d shards found but %d required", len(resp.ShardIdLocations), erasure_coding.DataShardsCount)
		}

		ecVolume.ShardLocationsLock.Lock()
		for _, shardIdLocations := range resp.ShardIdLocations {
			shardId := erasure_coding.ShardId(shardIdLocations.ShardId)
			delete(ecVolume.ShardLocations, shardId)
			for _, loc := range shardIdLocations.Locations {
				ecVolume.ShardLocations[shardId] = append(ecVolume.ShardLocations[shardId], pb.NewServerAddressFromLocation(loc))
			}
		}
		ecVolume.ShardLocationsRefreshTime = time.Now()
		ecVolume.ShardLocationsLock.Unlock()

		return nil
	})
	return
}

func (s *Store) readLocalEcShardInterval(ecVolume *erasure_coding.EcVolume, shardId erasure_coding.ShardId, buf []byte, offset int64) error {
	// Resolve the shard together with the EcVolume on the disk that owns it; the
	// shard may live on a sibling disk of this server.
	ownerVolume, shard, found := s.FindEcVolumeWithShard(ecVolume.VolumeId, shardId)
	if !found {
		return fmt.Errorf("shard %d for volume %d: %w", shardId, ecVolume.VolumeId, errShardNotLocal)
	}
	// Skip a local shard whose identity doesn't match the caller's index, so the
	// read recovers from the correct generation. Lenient only when the caller has
	// no identity (pre-upgrade): a known caller must not accept an unstamped local
	// shard, which would serve a stale pre-upgrade generation.
	if ecVolume.EncodeTsNs != 0 && ecVolume.EncodeTsNs != ownerVolume.EncodeTsNs {
		glog.V(1).Infof("skip local ec shard %d.%d from a different encode run: caller EncodeTsNs %d, local %d", ecVolume.VolumeId, shardId, ecVolume.EncodeTsNs, ownerVolume.EncodeTsNs)
		return fmt.Errorf("shard %d for volume %d: %w", shardId, ecVolume.VolumeId, errShardNotLocal)
	}

	readBytes, err := shard.ReadAt(buf, offset)
	if err != nil {
		return fmt.Errorf("failed to read local EC shard %d for volume %d: %v", shardId, ecVolume.VolumeId, err)
	}
	if got, want := readBytes, len(buf); got != want {
		return fmt.Errorf("expected %d bytes for local EC shard %d on volume %d, got %d", want, shardId, ecVolume.VolumeId, got)
	}

	return nil
}

func (s *Store) readRemoteEcShardInterval(sourceDataNodes []pb.ServerAddress, needleId types.NeedleId, vid needle.VolumeId, shardId erasure_coding.ShardId, buf []byte, offset int64, expectedEncodeTsNs int64) (n int, is_deleted bool, err error) {

	if len(sourceDataNodes) == 0 {
		return 0, false, fmt.Errorf("failed to find ec shard %d.%d", vid, shardId)
	}

	for _, sourceDataNode := range sourceDataNodes {
		glog.V(3).Infof("read remote ec shard %d.%d from %s", vid, shardId, sourceDataNode)
		n, is_deleted, err = s.doReadRemoteEcShardInterval(sourceDataNode, needleId, vid, shardId, buf, offset, expectedEncodeTsNs)
		if err == nil {
			return
		}
		glog.V(1).Infof("read remote ec shard %d.%d from %s: %v", vid, shardId, sourceDataNode, err)
	}

	return
}

func (s *Store) doReadRemoteEcShardInterval(sourceDataNode pb.ServerAddress, needleId types.NeedleId, vid needle.VolumeId, shardId erasure_coding.ShardId, buf []byte, offset int64, expectedEncodeTsNs int64) (n int, is_deleted bool, err error) {

	err = operation.WithVolumeServerClient(false, sourceDataNode, s.grpcDialOption, func(client volume_server_pb.VolumeServerClient) error {

		// copy data slice
		shardReadClient, err := client.VolumeEcShardRead(context.Background(), &volume_server_pb.VolumeEcShardReadRequest{
			VolumeId:   uint32(vid),
			ShardId:    uint32(shardId),
			Offset:     offset,
			Size:       int64(len(buf)),
			FileKey:    uint64(needleId),
			EncodeTsNs: expectedEncodeTsNs,
		})
		if err != nil {
			return fmt.Errorf("failed to start reading ec shard %d.%d from %s: %v", vid, shardId, sourceDataNode, err)
		}

		for {
			resp, receiveErr := shardReadClient.Recv()
			if receiveErr == io.EOF {
				break
			}
			if receiveErr != nil {
				return fmt.Errorf("receiving ec shard %d.%d from %s: %v", vid, shardId, sourceDataNode, receiveErr)
			}
			// Validate the served shard's identity client-side, so the guard holds
			// even against a pre-upgrade server that ignored the request field (it
			// returns 0). A mismatch fails the read; the caller recovers from parity.
			if expectedEncodeTsNs != 0 && resp.EncodeTsNs != expectedEncodeTsNs {
				return fmt.Errorf("ec shard %d.%d from %s belongs to a different encode run (want %d, got %d)", vid, shardId, sourceDataNode, expectedEncodeTsNs, resp.EncodeTsNs)
			}
			if resp.IsDeleted {
				is_deleted = true
			}
			copy(buf[n:n+len(resp.Data)], resp.Data)
			n += len(resp.Data)
		}

		return nil
	})
	if err != nil {
		return 0, is_deleted, fmt.Errorf("read ec shard %d.%d from %s: %v", vid, shardId, sourceDataNode, err)
	}

	// A non-deleted interval must arrive whole: the server stamps EncodeTsNs only
	// on chunks that carry bytes, so a short or empty stream (e.g. immediate EOF
	// from a pre-upgrade or stale server) leaves the buffer partly zero-filled and
	// unvalidated. Reject it so the caller recovers from parity. The is_deleted
	// short-circuit legitimately returns n=0 with no data and is exempt, matching
	// readLocalEcShardInterval's got==len(buf) rule for the local path.
	if !is_deleted && n != len(buf) {
		return n, is_deleted, fmt.Errorf("short read ec shard %d.%d from %s: got %d want %d", vid, shardId, sourceDataNode, n, len(buf))
	}

	return
}

func (s *Store) recoverOneRemoteEcShardInterval(needleId types.NeedleId, ecVolume *erasure_coding.EcVolume, shardIdToRecover erasure_coding.ShardId, buf []byte, offset int64) (n int, is_deleted bool, err error) {
	glog.V(3).Infof("recover ec shard %d.%d from other locations", ecVolume.VolumeId, shardIdToRecover)

	enc, err := reedsolomon.New(erasure_coding.DataShardsCount, erasure_coding.ParityShardsCount)
	if err != nil {
		return 0, false, fmt.Errorf("failed to create encoder: %w", err)
	}

	// Use MaxShardCount to support custom EC ratios up to 32 shards
	bufs := make([][]byte, erasure_coding.MaxShardCount)

	var wg sync.WaitGroup
	ecVolume.ShardLocationsLock.RLock()
	for shardId, locations := range ecVolume.ShardLocations {

		// skip current shard or empty shard
		if shardId == shardIdToRecover {
			continue
		}
		if len(locations) == 0 {
			glog.V(3).Infof("readRemoteEcShardInterval missing %d.%d from %+v", ecVolume.VolumeId, shardId, locations)
			continue
		}

		// read from remote locations
		wg.Add(1)
		go func(shardId erasure_coding.ShardId, locations []pb.ServerAddress) {
			defer wg.Done()
			data := make([]byte, len(buf))
			nRead, isDeleted, readErr := s.readRemoteEcShardInterval(locations, needleId, ecVolume.VolumeId, shardId, data, offset, ecVolume.EncodeTsNs)
			if readErr != nil {
				glog.V(3).Infof("recover: readRemoteEcShardInterval %d.%d %d bytes from %+v: %v", ecVolume.VolumeId, shardId, nRead, locations, readErr)
				forgetShardId(ecVolume, shardId)
			}
			if isDeleted {
				is_deleted = true
			}
			if nRead == len(buf) {
				bufs[shardId] = data
			}
		}(shardId, locations)
	}
	ecVolume.ShardLocationsLock.RUnlock()

	wg.Wait()

	// Count and log available shards for diagnostics
	availableShards := make([]erasure_coding.ShardId, 0, erasure_coding.TotalShardsCount)
	missingShards := make([]erasure_coding.ShardId, 0, erasure_coding.ParityShardsCount+1)
	for shardId := 0; shardId < erasure_coding.TotalShardsCount; shardId++ {
		if bufs[shardId] != nil {
			availableShards = append(availableShards, erasure_coding.ShardId(shardId))
		} else {
			missingShards = append(missingShards, erasure_coding.ShardId(shardId))
		}
	}

	glog.V(3).Infof("recover ec shard %d.%d: %d shards available %v, %d missing %v",
		ecVolume.VolumeId, shardIdToRecover,
		len(availableShards), availableShards,
		len(missingShards), missingShards)

	if len(availableShards) < erasure_coding.DataShardsCount {
		return 0, false, fmt.Errorf("cannot recover shard %d.%d: only %d shards available %v, need at least %d (missing: %v)",
			ecVolume.VolumeId, shardIdToRecover,
			len(availableShards), availableShards,
			erasure_coding.DataShardsCount, missingShards)
	}

	if err = enc.ReconstructData(bufs[:erasure_coding.TotalShardsCount]); err != nil {
		return 0, false, fmt.Errorf("failed to reconstruct data for shard %d.%d with %d available shards %v: %w",
			ecVolume.VolumeId, shardIdToRecover, len(availableShards), availableShards, err)
	}
	glog.V(4).Infof("recovered ec shard %d.%d from other locations", ecVolume.VolumeId, shardIdToRecover)

	copy(buf, bufs[shardIdToRecover])

	return len(buf), is_deleted, nil
}

func (s *Store) EcVolumes() (ecVolumes []*erasure_coding.EcVolume) {
	for _, location := range s.Locations {
		location.ecVolumesLock.RLock()
		for _, v := range location.ecVolumes {
			ecVolumes = append(ecVolumes, v)
		}
		location.ecVolumesLock.RUnlock()
	}
	slices.SortFunc(ecVolumes, func(a, b *erasure_coding.EcVolume) int {
		return int(a.VolumeId) - int(b.VolumeId)
	})
	return ecVolumes
}