Files
seaweedfs/weed/replication/replicator_test.go
T
Peter Dodd 5f1a742726 fix: return NoSuchKey and drop stale entries when a remote-mounted object is gone from the remote (#11680)
A remote-only entry whose object was deleted from the remote storage
outside the filer answered GET with 500 and stayed in the filer. The
remote's not-found was lost on the way: the backends' ReadFile returned
it as an untyped error, so FetchAndWriteNeedle failed with codes.Unknown
and nothing downstream could tell it from any other failure.

- remote_storage: GCS, S3 (NoSuchKey) and Azure (BlobNotFound) reads
  return ErrRemoteObjectNotFound. GCS reports a missing bucket the same
  way as a missing object, so it confirms the bucket with a listing.
- volume server and filer: the not-found crosses gRPC as codes.NotFound
  carrying the sentinel's text, and the filer's cache RPC returns
  codes.NotFound, which the S3 gateway already maps to NoSuchKey.
- s3api: the origin fallback answers NoSuchKey on a confirmed not-found.
- filer: a confirmed not-found removes the stale entry, so the lazy
  remote-metadata cache converges on the remote. Only remote-only files
  outside .versions and without an active object lock are removed, only
  if unchanged since the fetch (checked on the object's write owner,
  under the lock S3 object writes take), and with a metadata-only
  delete: the filer skips its inline remote delete, and the delete
  events' entries carry a marker that makes filer.remote.sync and
  filer.remote.gateway skip their remote delete, while filer.sync still
  replicates it. The replicated DeleteEntryRequest carries
  keep_remote_object, so the destination's delete events are marked too.
  The store drops the marker from every write, so clients cannot plant
  it.
2026-10-10 11:00:04 +08:00

447 lines
13 KiB
Go

package replication
import (
"context"
"testing"
"github.com/seaweedfs/seaweedfs/weed/filer"
"github.com/seaweedfs/seaweedfs/weed/pb/filer_pb"
"github.com/seaweedfs/seaweedfs/weed/replication/sink"
"github.com/seaweedfs/seaweedfs/weed/replication/source"
"github.com/seaweedfs/seaweedfs/weed/util"
)
var _ sink.ReplicationSink = (*recordingSink)(nil)
type deleteCall struct {
key string
isDirectory bool
}
type createCall struct {
key string
}
type updateCall struct {
key string
newParentPath string
}
type recordingSink struct {
name string
sinkToDirectory string
incremental bool
updateFoundExisting bool
deleteCalls []deleteCall
createCalls []createCall
updateCalls []updateCall
// ordered records the sink method names in call order so tests can assert
// create-before-delete sequencing.
ordered []string
}
func (s *recordingSink) GetName() string {
return s.name
}
func (s *recordingSink) Initialize(util.Configuration, string) error {
return nil
}
func (s *recordingSink) DeleteEntry(key string, isDirectory, deleteIncludeChunks bool, signatures []int32) error {
s.deleteCalls = append(s.deleteCalls, deleteCall{key: key, isDirectory: isDirectory})
s.ordered = append(s.ordered, "delete")
return nil
}
func (s *recordingSink) CreateEntry(key string, entry *filer_pb.Entry, signatures []int32) error {
s.createCalls = append(s.createCalls, createCall{key: key})
s.ordered = append(s.ordered, "create")
return nil
}
func (s *recordingSink) UpdateEntry(key string, oldEntry *filer_pb.Entry, newParentPath string, newEntry *filer_pb.Entry, deleteIncludeChunks bool, signatures []int32) (bool, error) {
s.updateCalls = append(s.updateCalls, updateCall{key: key, newParentPath: newParentPath})
s.ordered = append(s.ordered, "update")
return s.updateFoundExisting, nil
}
func (s *recordingSink) GetSinkToDirectory() string {
return s.sinkToDirectory
}
func (s *recordingSink) GetDestinationIdentity() string {
return s.sinkToDirectory
}
func (s *recordingSink) SetSourceFiler(*source.FilerSource) {}
func (s *recordingSink) IsIncremental() bool {
return s.incremental
}
type moveCall struct {
oldKey string
newKey string
}
// movingSink is a recordingSink that also implements sink.EntryMover, modeling
// a sink (like the filer) with a native atomic move.
type movingSink struct {
*recordingSink
moveCalls []moveCall
}
var _ sink.EntryMover = (*movingSink)(nil)
func (s *movingSink) MoveEntry(oldKey, newKey string, newEntry *filer_pb.Entry, signatures []int32) error {
s.moveCalls = append(s.moveCalls, moveCall{oldKey: oldKey, newKey: newKey})
s.ordered = append(s.ordered, "move")
return nil
}
func TestReplicateRenameUsesTargetKeyForNonFilerSink(t *testing.T) {
s := &recordingSink{name: "local", sinkToDirectory: "/dest"}
r := &Replicator{
sink: s,
source: &source.FilerSource{Dir: "/source"},
}
err := r.Replicate(context.Background(), "/source/old/file.txt", &filer_pb.EventNotification{
OldEntry: &filer_pb.Entry{
Name: "file.txt",
Attributes: &filer_pb.FuseAttributes{
Mtime: 123,
},
},
NewEntry: &filer_pb.Entry{
Name: "renamed.txt",
Attributes: &filer_pb.FuseAttributes{
Mtime: 123,
},
},
NewParentPath: "/source/new",
})
if err != nil {
t.Fatalf("Replicate rename: %v", err)
}
if len(s.updateCalls) != 0 {
t.Fatalf("expected rename to bypass UpdateEntry, got %d calls", len(s.updateCalls))
}
if len(s.createCalls) != 1 || s.createCalls[0].key != "/dest/new/renamed.txt" {
t.Fatalf("create calls = %+v, want target sink key", s.createCalls)
}
if len(s.deleteCalls) != 1 || s.deleteCalls[0].key != "/dest/old/file.txt" {
t.Fatalf("delete calls = %+v, want old sink key", s.deleteCalls)
}
if got, want := s.ordered, []string{"create", "delete"}; !equalStrings(got, want) {
t.Fatalf("call order = %v, want create before delete %v", got, want)
}
}
// A real move to a filer sink also goes through create-then-delete (the
// filer-sink special-case that previously routed renames to UpdateEntry is
// gone): UpdateEntry cannot move an entry across paths.
// A sink without a native move falls back to create-then-delete for a rename.
func TestReplicateRenameWithoutMoverUsesCreateThenDelete(t *testing.T) {
s := &recordingSink{
name: "s3",
sinkToDirectory: "/dest",
updateFoundExisting: true,
}
r := &Replicator{
sink: s,
source: &source.FilerSource{Dir: "/source"},
}
err := r.Replicate(context.Background(), "/source/old/file.txt", &filer_pb.EventNotification{
OldEntry: &filer_pb.Entry{
Name: "file.txt",
Attributes: &filer_pb.FuseAttributes{
Mtime: 123,
},
},
NewEntry: &filer_pb.Entry{
Name: "renamed.txt",
Attributes: &filer_pb.FuseAttributes{
Mtime: 123,
},
},
NewParentPath: "/source/new",
})
if err != nil {
t.Fatalf("Replicate rename: %v", err)
}
if len(s.updateCalls) != 0 {
t.Fatalf("expected rename to bypass UpdateEntry, got %d calls", len(s.updateCalls))
}
if len(s.createCalls) != 1 || s.createCalls[0].key != "/dest/new/renamed.txt" {
t.Fatalf("create calls = %+v, want target sink key", s.createCalls)
}
if len(s.deleteCalls) != 1 || s.deleteCalls[0].key != "/dest/old/file.txt" {
t.Fatalf("delete calls = %+v, want old sink key", s.deleteCalls)
}
if got, want := s.ordered, []string{"create", "delete"}; !equalStrings(got, want) {
t.Fatalf("call order = %v, want create before delete %v", got, want)
}
}
// A sink with a native move (the filer) relocates a rename via MoveEntry — never
// create-then-delete, so a failed copy can't delete the only valid destination.
func TestReplicateRenameUsesMoveEntryWhenSupported(t *testing.T) {
s := &movingSink{recordingSink: &recordingSink{name: "filer", sinkToDirectory: "/dest"}}
r := &Replicator{
sink: s,
source: &source.FilerSource{Dir: "/source"},
}
err := r.Replicate(context.Background(), "/source/old/file.txt", &filer_pb.EventNotification{
OldEntry: &filer_pb.Entry{Name: "file.txt"},
NewEntry: &filer_pb.Entry{Name: "renamed.txt"},
NewParentPath: "/source/new",
})
if err != nil {
t.Fatalf("Replicate rename: %v", err)
}
if len(s.moveCalls) != 1 || s.moveCalls[0].oldKey != "/dest/old/file.txt" || s.moveCalls[0].newKey != "/dest/new/renamed.txt" {
t.Fatalf("move calls = %+v, want one move /dest/old/file.txt => /dest/new/renamed.txt", s.moveCalls)
}
if len(s.createCalls) != 0 || len(s.deleteCalls) != 0 || len(s.updateCalls) != 0 {
t.Fatalf("native move must not create/delete/update: creates=%v deletes=%v updates=%v", s.createCalls, s.deleteCalls, s.updateCalls)
}
}
// An in-place update (same parent + same name, so oldSinkKey == newSinkKey)
// still routes to UpdateEntry rather than create-then-delete.
func TestReplicateInPlaceUpdateUsesUpdateEntry(t *testing.T) {
s := &recordingSink{
name: "filer",
sinkToDirectory: "/dest",
updateFoundExisting: true,
}
r := &Replicator{
sink: s,
source: &source.FilerSource{Dir: "/source"},
}
err := r.Replicate(context.Background(), "/source/dir/file.txt", &filer_pb.EventNotification{
OldEntry: &filer_pb.Entry{
Name: "file.txt",
Attributes: &filer_pb.FuseAttributes{
Mtime: 123,
},
},
NewEntry: &filer_pb.Entry{
Name: "file.txt",
Attributes: &filer_pb.FuseAttributes{
Mtime: 456,
},
},
NewParentPath: "/source/dir",
})
if err != nil {
t.Fatalf("Replicate in-place update: %v", err)
}
if len(s.updateCalls) != 1 {
t.Fatalf("update calls = %d, want 1", len(s.updateCalls))
}
if s.updateCalls[0].key != "/dest/dir/file.txt" {
t.Fatalf("update key = %q, want /dest/dir/file.txt", s.updateCalls[0].key)
}
if s.updateCalls[0].newParentPath != "/dest/dir" {
t.Fatalf("update newParentPath = %q, want /dest/dir", s.updateCalls[0].newParentPath)
}
if len(s.deleteCalls) != 0 || len(s.createCalls) != 0 {
t.Fatalf("unexpected delete/create calls: deletes=%+v creates=%+v", s.deleteCalls, s.createCalls)
}
}
// When the in-place update finds no existing entry to update, fall back to
// delete-then-create at the same key.
func TestReplicateInPlaceUpdateFallbackCreates(t *testing.T) {
s := &recordingSink{
name: "filer",
sinkToDirectory: "/dest",
updateFoundExisting: false,
}
r := &Replicator{
sink: s,
source: &source.FilerSource{Dir: "/source"},
}
err := r.Replicate(context.Background(), "/source/dir/file.txt", &filer_pb.EventNotification{
OldEntry: &filer_pb.Entry{
Name: "file.txt",
Attributes: &filer_pb.FuseAttributes{
Mtime: 123,
},
},
NewEntry: &filer_pb.Entry{
Name: "file.txt",
Attributes: &filer_pb.FuseAttributes{
Mtime: 456,
},
},
NewParentPath: "/source/dir",
})
if err != nil {
t.Fatalf("Replicate in-place update fallback: %v", err)
}
if len(s.updateCalls) != 1 {
t.Fatalf("update calls = %d, want 1", len(s.updateCalls))
}
if len(s.deleteCalls) != 1 || s.deleteCalls[0].key != "/dest/dir/file.txt" {
t.Fatalf("delete calls = %+v, want same sink key", s.deleteCalls)
}
if len(s.createCalls) != 1 || s.createCalls[0].key != "/dest/dir/file.txt" {
t.Fatalf("create calls = %+v, want same sink key", s.createCalls)
}
if got, want := s.ordered, []string{"update", "delete", "create"}; !equalStrings(got, want) {
t.Fatalf("call order = %v, want %v", got, want)
}
}
func equalStrings(a, b []string) bool {
if len(a) != len(b) {
return false
}
for i := range a {
if a[i] != b[i] {
return false
}
}
return true
}
func TestPathIsEqualOrUnderUsesDirectoryBoundaries(t *testing.T) {
tests := []struct {
name string
candidate string
other string
expected bool
}{
{name: "equal", candidate: "/foo", other: "/foo", expected: true},
{name: "descendant", candidate: "/foo/bar", other: "/foo", expected: true},
{name: "sibling prefix", candidate: "/foobar/bar", other: "/foo", expected: false},
{name: "root", candidate: "/foo/bar", other: "/", expected: true},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
if got := util.IsEqualOrUnder(tt.candidate, tt.other); got != tt.expected {
t.Fatalf("IsEqualOrUnder(%q, %q) = %v, want %v", tt.candidate, tt.other, got, tt.expected)
}
})
}
}
func TestReplicateRenameOutToSiblingPrefixBecomesDelete(t *testing.T) {
s := &recordingSink{name: "local", sinkToDirectory: "/dest"}
r := &Replicator{
sink: s,
source: &source.FilerSource{Dir: "/foo"},
}
err := r.Replicate(context.Background(), "/foo/old/file.txt", &filer_pb.EventNotification{
OldEntry: &filer_pb.Entry{
Name: "file.txt",
Attributes: &filer_pb.FuseAttributes{
Mtime: 123,
},
},
NewEntry: &filer_pb.Entry{
Name: "file.txt",
Attributes: &filer_pb.FuseAttributes{
Mtime: 123,
},
},
NewParentPath: "/foobar/new",
})
if err != nil {
t.Fatalf("Replicate rename out to sibling prefix: %v", err)
}
if len(s.deleteCalls) != 1 || s.deleteCalls[0].key != "/dest/old/file.txt" {
t.Fatalf("delete calls = %+v, want old sink key", s.deleteCalls)
}
if len(s.createCalls) != 0 || len(s.updateCalls) != 0 {
t.Fatalf("unexpected create/update calls: creates=%+v updates=%+v", s.createCalls, s.updateCalls)
}
}
func TestReplicateRenameFromExcludedDirBecomesCreate(t *testing.T) {
s := &recordingSink{name: "local", sinkToDirectory: "/dest"}
r := &Replicator{
sink: s,
source: &source.FilerSource{Dir: "/foo"},
excludeDirs: []string{"/foo/excluded"},
}
err := r.Replicate(context.Background(), "/foo/excluded/file.txt", &filer_pb.EventNotification{
OldEntry: &filer_pb.Entry{
Name: "file.txt",
Attributes: &filer_pb.FuseAttributes{
Mtime: 123,
},
},
NewEntry: &filer_pb.Entry{
Name: "file.txt",
Attributes: &filer_pb.FuseAttributes{
Mtime: 123,
},
},
NewParentPath: "/foo/live",
})
if err != nil {
t.Fatalf("Replicate rename from excluded dir: %v", err)
}
if len(s.createCalls) != 1 || s.createCalls[0].key != "/dest/live/file.txt" {
t.Fatalf("create calls = %+v, want target sink key", s.createCalls)
}
if len(s.deleteCalls) != 0 || len(s.updateCalls) != 0 {
t.Fatalf("unexpected delete/update calls: deletes=%+v updates=%+v", s.deleteCalls, s.updateCalls)
}
}
type keepingSink struct {
*recordingSink
keepCalls []deleteCall
}
var _ sink.MetadataOnlyDeleter = (*keepingSink)(nil)
func (s *keepingSink) DeleteEntryKeepingRemoteObject(key string, isDirectory, deleteIncludeChunks bool, signatures []int32) error {
s.keepCalls = append(s.keepCalls, deleteCall{key: key, isDirectory: isDirectory})
return nil
}
func TestReplicateMetadataOnlyDeleteKeepsTheRemoteObject(t *testing.T) {
s := &keepingSink{recordingSink: &recordingSink{name: "filer", sinkToDirectory: "/dest"}}
r := &Replicator{sink: s, source: &source.FilerSource{Dir: "/source"}}
deleteOf := func(name string, extended map[string][]byte) *filer_pb.EventNotification {
return &filer_pb.EventNotification{OldEntry: &filer_pb.Entry{Name: name, Attributes: &filer_pb.FuseAttributes{}, Extended: extended}}
}
if err := r.Replicate(context.Background(), "/source/dir/kept.bin", deleteOf("kept.bin", map[string][]byte{filer.ExtKeepRemoteObjectKey: []byte("true")})); err != nil {
t.Fatal(err)
}
if err := r.Replicate(context.Background(), "/source/dir/gone.bin", deleteOf("gone.bin", nil)); err != nil {
t.Fatal(err)
}
if len(s.keepCalls) != 1 || s.keepCalls[0].key != "/dest/dir/kept.bin" {
t.Fatalf("metadata-only deletes = %+v, want /dest/dir/kept.bin", s.keepCalls)
}
if len(s.deleteCalls) != 1 || s.deleteCalls[0].key != "/dest/dir/gone.bin" {
t.Fatalf("ordinary deletes = %+v, want /dest/dir/gone.bin", s.deleteCalls)
}
}