ec.rebuild: remount shards the rebuild replaced, keep regenerated copies (#12193)

* volume: remount ec shards the rebuild replaced under a mounted handle

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* ec.rebuild: keep copied shards the rebuild regenerated in place

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* volume: reopen regenerated ec shards in place instead of remounting

UnmountEcShards and MountEcShards send deltas through separate buffered
channels, so a remount can reach the master as add-then-delete and drop
the repaired shard from lookups until the next full EC heartbeat.
ReopenEcShard swaps the mounted shard's descriptor under the volume lock
instead: registration, disk type, and heartbeat state never flip, and
readers holding the old shard degrade to a clean error the parity path
already recovers from.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* rust volume: reopen rebuilt ec shards in place, mirroring Go

volume_ec_shards_rebuild recreates a shard file under a mounted handle
whose descriptor still serves the detached inode (a mounted shard whose
file was deleted under it). DiskLocation::reopen_ec_shard swaps the
mounted descriptor for a fresh open, matching Go's ReopenEcShard — the
registration and heartbeat view never flip.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* volume: fail ec rebuild when a regenerated shard cannot be reopened

* rust volume: fail ec shards rebuild when a rebuilt shard cannot be reopened

* volume: repoint mounted ec shards after rebuild across disks

A mounted shard whose file was deleted under the mount is "missing" to
the rebuild: the regenerated file lands on the rebuild location while
the handle keeps serving the detached inode on another disk. Reopening
the mounted disk then failed the whole RPC, so the caller never saw the
rebuilt ids and its deferred cleanup deleted the regenerated copies.

Repoint every mounted handle of the rebuilt volume instead of only this
run's outputs: reopen in place when the rebuilt file is on the mounted
disk, otherwise drop the stale handle and remount the shard from
whichever local disk holds the file via a delta-free load split out of
MountEcShards — a same-server move, so the master's registration never
flips. A shard that can be neither reopened nor remounted is left
unmounted rather than serving replaced bytes, and the repoint step no
longer fails the RPC, so RebuiltShardIds always reaches the caller.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* rust volume: repoint mounted ec shards after rebuild across disks

Mirror the Go repoint loop: iterate every mounted shard of the rebuilt
volume, reopen in place when the rebuilt file is on the mounted disk,
and otherwise drop the stale handle and remount the shard from
whichever local disk holds it. The repoint no longer fails the RPC, so
rebuilt_shard_ids still reaches the caller and its copy cleanup keeps
the regenerated shards.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* rust volume: still rebuild ecx and repoint mounts when no shard is missing

* master: keep a node in an ec shard's lookup while a disk still holds it

A rebuild that relocates a mounted shard between a node's disks reports
the move as a delete on the source disk plus an add on the destination.
Registering first is a no-op because shard lookups key on the node, then
the unregister dropped the node outright — and steady-state heartbeats
never restore it, so readers and later rebuilds lose the repaired copy.
Skip the delete while another disk on the node still holds the shard.

UnregisterDataNode now clears the node's own ec records before dropping
its registrations, so the retained-shard check cannot skip the teardown,
and each unregister gets a ShardsInfo copy since DeleteEcShard drains
the shared one.

* rust volume: fold the journal on a mounted index instead of rewriting it

The rebuild handler regenerated .ecx through File::create even while the
volume was mounted, truncating the index under readers that search the
open descriptor with its old size — a live lookup could hit EOF mid-scan.
Go folds the deletion journal in place on a mounted volume; do the same
through rebuild_ecx_from_journal and reserve the full regen for the
unmounted path, where nothing holds the file open to interrupt.

* volume, rust volume: report rebuilt shards on late failures, fix rust corrupt discovery

Two review findings the repoint rework left open:

- A RebuildEcxFile failure still returns a plain error after shards were
  regenerated, so the caller discards the response holding the rebuilt
  ids and the deferred cleanup deletes copies the rebuild regenerated
  in place. The Go handler now packs the ids into the gRPC status
  details and the shell pulls them back out on error; the Rust handler
  carries them in grpc-status-details-bin the same way.
- The Rust rebuild still counted a sidecar-corrupt shard as present, so
  it fed bad bytes into reconstruction and never regenerated the shard.
  Discovery now loads the generation-0 .ecsum the way Go does, verifies
  every present shard, and reclassifies mismatches (and zero-size
  residue) as rebuild targets written to a .rebuilding temp file,
  verified, and atomically renamed over the original only after every
  output passes the sidecar check. Honors unsafe_ignore_sidecar, added
  to VolumeEcShardsRebuildRequest to match the Go proto.

* volume: repoint sibling journals sharing a folded .ecj after rebuild

RebuildEcxFile unlinks the deletion journal and the mounted path only
reopens its own handle — a sibling mount journaling into the same file
keeps appending to the detached inode, and those deletes never reach a
path-based reader again. Reopen every mounted volume sharing the index
base before leaving the rebuild.

* rust volume: repoint sibling journals sharing a folded .ecj after rebuild

rebuild_ecx_from_journal replaces the journal and reopens only its own
handle; mounts on sibling disks sharing that .ecj keep the detached
handle, so their next appends are lost at restart. Adopt the folded
journal on every mounted runtime sharing the path before releasing the
store lock, clearing their in-memory set — every journaled id is now an
.ecx tombstone.

* ec: swap the deletion journal handle only after the new file opens

A failed reopen left ecjFile nil, so every later delete returned
"ec volume closed". Open first, then close and swap — mirroring the
Rust adopt_folded_journal ordering.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* ec.rebuild: fail the rebuild when a sibling journal cannot be repointed

A warning left the sibling mount journaling to a detached inode while
the RPC reported success. Propagate the error like the Rust handler's
adopt_folded_journal map_err.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* ec: keep deletes loud when the journal reopen cannot open the replacement

Reverts the swap-on-success ordering: a detached handle kept on failure
lets deletes append and sync to the removed inode, then vanish at
restart. Closing first also frees the descriptor before the open.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* rust volume: drop the folded journal handle before reopening it

adopt_folded_journal and the post-fold recreate now clear ecj_file
before opening the replacement, so a failed reopen leaves deletes
erroring instead of appending to the detached inode. Mirrors Go's
ReopenDeletionJournal.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* ec: drop the cached journal size when its reopen fails

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

---------

Co-authored-by: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

weed/server/volume_grpc_ec_rebuild_reopen_test.go

@@ -0,0 +1,185 @@
+package weed_server
+
+import (
+	"bytes"
+	"context"
+	"crypto/rand"
+	"os"
+	"path/filepath"
+	"testing"
+
+	"github.com/seaweedfs/seaweedfs/weed/pb/volume_server_pb"
+	"github.com/seaweedfs/seaweedfs/weed/stats"
+	"github.com/seaweedfs/seaweedfs/weed/storage"
+	"github.com/seaweedfs/seaweedfs/weed/storage/erasure_coding"
+	"github.com/seaweedfs/seaweedfs/weed/storage/needle"
+	"github.com/seaweedfs/seaweedfs/weed/storage/types"
+	"github.com/seaweedfs/seaweedfs/weed/util"
+	"github.com/stretchr/testify/require"
+)
+
+// A corrupt shard that stays mounted while VolumeEcShardsRebuild regenerates it
+// in place must stop serving the pre-repair bytes: the handler reopens the
+// mounted shard so reads hit the regenerated file, not the detached inode.
+// Regression test for ec.rebuild leaving a replaced shard file open until
+// restart.
+func TestVolumeEcShardsRebuild_ReopensMountedCorruptShard(t *testing.T) {
+	if !erasure_coding.BitrotProtectionEnabled {
+		t.Skip("bitrot protection is compiled out")
+	}
+	dataDir := filepath.Join(t.TempDir(), "data")
+	require.NoError(t, os.MkdirAll(dataDir, 0o755))
+	store := storage.NewStore(nil, "localhost", 8080, 18080, "http://localhost:8080", "store-id",
+		[]string{dataDir}, []int32{100}, []util.MinFreeSpace{{}}, "",
+		storage.NeedleMapInMemory, []types.DiskType{types.HardDriveType}, nil, 3, stats.DefaultDiskIOProbeConfig())
+	done := make(chan struct{})
+	go func() {
+		for {
+			select {
+			case <-store.NewVolumesChan:
+			case <-store.NewEcShardsChan:
+			case <-store.DeletedVolumesChan:
+			case <-store.DeletedEcShardsChan:
+			case <-store.StateUpdateChan:
+			case <-done:
+				return
+			}
+		}
+	}()
+	t.Cleanup(func() {
+		store.Close()
+		close(done)
+	})
+	st, err := storage.NewState(t.TempDir())
+	require.NoError(t, err)
+	store.State = st
+	vs := &VolumeServer{store: store}
+
+	collection := "col"
+	vid := needle.VolumeId(42)
+	base := erasure_coding.EcShardFileName(collection, dataDir, int(vid))
+
+	dat := make([]byte, 7000)
+	_, err = rand.Read(dat)
+	require.NoError(t, err)
+	require.NoError(t, os.WriteFile(base+".dat", dat, 0o644))
+	prot, err := erasure_coding.WriteEcFiles(base, nil)
+	require.NoError(t, err)
+	require.NoError(t, erasure_coding.SaveBitrotSidecar(erasure_coding.BitrotSidecarPath(base, 0), prot))
+	require.NoError(t, os.WriteFile(base+".ecx", make([]byte, 16), 0o644))
+
+	require.NoError(t, store.MountEcShards(collection, vid, 0, ""))
+	readShard := func() []byte {
+		shard, found := store.Locations[0].FindEcShard(vid, 0)
+		require.True(t, found)
+		buf := make([]byte, shard.Size())
+		_, err := shard.ReadAt(buf, 0)
+		require.NoError(t, err)
+		return buf
+	}
+	good := readShard()
+
+	// Corrupt the file in place: the mounted descriptor now serves bad bytes.
+	corrupt := bytes.Repeat([]byte{0xBB}, 1024)
+	f, err := os.OpenFile(base+erasure_coding.ToExt(0), os.O_WRONLY, 0)
+	require.NoError(t, err)
+	_, err = f.WriteAt(corrupt, 0)
+	require.NoError(t, err)
+	require.NoError(t, f.Close())
+	require.True(t, bytes.Equal(corrupt, readShard()[:len(corrupt)]))
+
+	_, err = vs.VolumeEcShardsRebuild(context.Background(), &volume_server_pb.VolumeEcShardsRebuildRequest{
+		VolumeId:   uint32(vid),
+		Collection: collection,
+	})
+	require.NoError(t, err)
+
+	require.True(t, bytes.Equal(good, readShard()),
+		"mounted shard should serve the regenerated file after rebuild")
+}
+
+// A shard whose file is deleted under the mount is "missing" to the rebuild:
+// it regenerates on the rebuild disk while the mounted handle still serves
+// the detached inode on another disk. Reopening the mounted disk cannot fix
+// that, so the handler moves the registration to the disk holding the rebuilt
+// file — and still reports the rebuilt ids so the caller's copy cleanup keeps
+// them.
+func TestVolumeEcShardsRebuild_RelocatesMountedShardAcrossDisks(t *testing.T) {
+	if !erasure_coding.BitrotProtectionEnabled {
+		t.Skip("bitrot protection is compiled out")
+	}
+	tempDir := t.TempDir()
+	dirA := filepath.Join(tempDir, "a")
+	dirB := filepath.Join(tempDir, "b")
+	require.NoError(t, os.MkdirAll(dirA, 0o755))
+	require.NoError(t, os.MkdirAll(dirB, 0o755))
+	store := storage.NewStore(nil, "localhost", 8080, 18080, "http://localhost:8080", "store-id",
+		[]string{dirA, dirB}, []int32{100, 100}, []util.MinFreeSpace{{}, {}}, "",
+		storage.NeedleMapInMemory, []types.DiskType{types.HardDriveType, types.HardDriveType}, nil, 3, stats.DefaultDiskIOProbeConfig())
+	done := make(chan struct{})
+	go func() {
+		for {
+			select {
+			case <-store.NewVolumesChan:
+			case <-store.NewEcShardsChan:
+			case <-store.DeletedVolumesChan:
+			case <-store.DeletedEcShardsChan:
+			case <-store.StateUpdateChan:
+			case <-done:
+				return
+			}
+		}
+	}()
+	t.Cleanup(func() {
+		store.Close()
+		close(done)
+	})
+	st, err := storage.NewState(t.TempDir())
+	require.NoError(t, err)
+	store.State = st
+	vs := &VolumeServer{store: store}
+
+	collection := "col"
+	vid := needle.VolumeId(42)
+	baseA := erasure_coding.EcShardFileName(collection, dirA, int(vid))
+	baseB := erasure_coding.EcShardFileName(collection, dirB, int(vid))
+	shardName := erasure_coding.ToExt(0)
+
+	dat := make([]byte, 7000)
+	_, err = rand.Read(dat)
+	require.NoError(t, err)
+	require.NoError(t, os.WriteFile(baseA+".dat", dat, 0o644))
+	prot, err := erasure_coding.WriteEcFiles(baseA, nil)
+	require.NoError(t, err)
+	require.NoError(t, erasure_coding.SaveBitrotSidecar(erasure_coding.BitrotSidecarPath(baseA, 0), prot))
+	require.NoError(t, os.WriteFile(baseA+".ecx", make([]byte, 16), 0o644))
+
+	original, err := os.ReadFile(baseA + shardName)
+	require.NoError(t, err)
+
+	// Mount shard 0 from disk B, then delete the file under the mount: the
+	// handle keeps serving the detached inode while the disk reports the
+	// shard missing.
+	require.NoError(t, os.Rename(baseA+shardName, baseB+shardName))
+	require.NoError(t, store.MountEcShards(collection, vid, 0, ""))
+	diskId, _, found := store.FindEcShard(vid, 0)
+	require.True(t, found)
+	require.Equal(t, uint32(1), diskId, "shard 0 should be mounted on disk B")
+	require.NoError(t, os.Remove(baseB+shardName))
+
+	resp, err := vs.VolumeEcShardsRebuild(context.Background(), &volume_server_pb.VolumeEcShardsRebuildRequest{
+		VolumeId:   uint32(vid),
+		Collection: collection,
+	})
+	require.NoError(t, err)
+	require.Equal(t, []uint32{0}, resp.RebuiltShardIds,
+		"a mounted shard whose file was rebuilt on another disk must still be reported rebuilt")
+
+	diskId, shard, found := store.FindEcShard(vid, 0)
+	require.True(t, found)
+	require.Equal(t, uint32(0), diskId, "rebuilt shard should be mounted on the disk holding its file")
+	buf := make([]byte, shard.Size())
+	_, err = shard.ReadAt(buf, 0)
+	require.NoError(t, err)
+	require.True(t, bytes.Equal(original, buf), "relocated shard should serve the regenerated file")
+}

weed/server/volume_grpc_erasure_coding.go

@@ -186,6 +186,19 @@ func recordEcRebuild(result string, d time.Duration) {
 	stats.VolumeServerECRebuildCounter.WithLabelValues(result).Inc()
 }
 
+// ecRebuildErrorWithIds reports a rebuild failure that happened after shards
+// were regenerated on disk. The error still carries their ids so the caller's
+// deferred cleanup keeps the copies it regenerated in place — to the shell an
+// unreported regenerated copy is just an orphan to delete, and a plain gRPC
+// error would drop the response holding them.
+func ecRebuildErrorWithIds(rebuiltShardIds []uint32, err error) error {
+	st := status.New(codes.Internal, err.Error())
+	if withIds, derr := st.WithDetails(&volume_server_pb.VolumeEcShardsRebuildResponse{RebuiltShardIds: rebuiltShardIds}); derr == nil {
+		st = withIds
+	}
+	return st.Err()
+}
+
 // VolumeEcShardsRebuild generates the any of the missing .ec00 ~ .ec13 files
 func (vs *VolumeServer) VolumeEcShardsRebuild(ctx context.Context, req *volume_server_pb.VolumeEcShardsRebuildRequest) (*volume_server_pb.VolumeEcShardsRebuildResponse, error) {
 	if err := vs.checkGrpcAdminAuth(ctx); err != nil {
@@ -280,13 +293,58 @@ func (vs *VolumeServer) VolumeEcShardsRebuild(ctx context.Context, req *volume_s
 			err = ev.ReopenDeletionJournal()
 		}
 		unlockJournal()
+		if err == nil {
+			// Mounts on sibling disks can journal into the same .ecj — the
+			// fold unlinked it, so repoint their handles too or their next
+			// append lands on the detached inode.
+			for _, location := range vs.store.Locations {
+				if location == rebuildLocation {
+					continue
+				}
+				if sibling, ok := location.FindEcVolume(needle.VolumeId(req.VolumeId)); ok && sibling.EcIndexBaseFileName() == indexBaseFileName {
+					unlockSibling := sibling.LockDeletionJournal()
+					serr := sibling.ReopenDeletionJournal()
+					unlockSibling()
+					if serr != nil {
+						err = fmt.Errorf("reopen shared ec journal on %s: %v", location.Directory, serr)
+						break
+					}
+				}
+			}
+		}
 		if err != nil {
 			recordEcRebuild("failure", time.Since(start))
-			return nil, fmt.Errorf("RebuildEcxFile %s: %v", indexBaseFileName, err)
+			return nil, ecRebuildErrorWithIds(rebuiltShardIds, fmt.Errorf("RebuildEcxFile %s: %v", indexBaseFileName, err))
 		}
 	} else if err := erasure_coding.RebuildEcxFile(indexBaseFileName); err != nil {
 		recordEcRebuild("failure", time.Since(start))
-		return nil, fmt.Errorf("RebuildEcxFile %s: %v", indexBaseFileName, err)
+		return nil, ecRebuildErrorWithIds(rebuiltShardIds, fmt.Errorf("RebuildEcxFile %s: %v", indexBaseFileName, err))
+	}
+
+	// A regenerated shard that was mounted keeps serving its old file through
+	// the open descriptor — the rename left the mounted handle on the detached
+	// inode. Repoint every mounted handle of this volume at the file now on
+	// disk, not just this run's outputs: a shard an earlier rebuild wrote on
+	// another disk is "present" this run and would never re-enter
+	// generatedShardIds, yet its mounted handle is still stale.
+	for _, location := range vs.store.Locations {
+		for _, shardId := range location.MountedEcShardIds(needle.VolumeId(req.VolumeId)) {
+			if _, rerr := location.ReopenEcShard(needle.VolumeId(req.VolumeId), shardId); rerr == nil {
+				continue
+			}
+			// The mounted disk lost the file and the rebuild wrote it
+			// elsewhere — or the file is unreadable in place. Drop the stale
+			// handle and remount from whichever local disk holds the rebuilt
+			// file; both steps emit no heartbeat delta, so the master's
+			// registration never flips. A shard that cannot be remounted is
+			// left unmounted — the honest state — rather than serving
+			// replaced bytes. Reporting the rebuilt ids on this path keeps
+			// the caller's cleanup from deleting them as strays.
+			location.UnloadEcShard(needle.VolumeId(req.VolumeId), shardId)
+			if _, _, lerr := vs.store.LoadEcShardAcrossDisks(req.Collection, needle.VolumeId(req.VolumeId), shardId); lerr != nil {
+				glog.Warningf("rebuilt ec shard %d.%d left unmounted: %v", req.VolumeId, shardId, lerr)
+			}
+		}
 	}
 
 	recordEcRebuild("success", time.Since(start))