fix: add more optimizations to putMetacacheObject() (#12916)

- avoid extra lookup for 'xl.meta' since we are
  definitely sure that it doesn't exist.

- use this in newMultipartUpload() as well

- also additionally do not write with O_DSYNC
  to avoid loading the drives, instead create
  'xl.meta' for listing operations without
  O_DSYNC since these are ephemeral objects.

- do the same with newMultipartUpload() since
  it gets synced when the PutObjectPart() is
  attempted, we do not need to tax newMultipartUpload()
  instead.

cmd/erasure-multipart.go

@@ -347,8 +347,9 @@ func (er erasureObjects) newMultipartUpload(ctx context.Context, bucket string,
 	// Fill all the necessary metadata.
 	// Update `xl.meta` content on each disks.
 	for index := range partsMetadata {
-		partsMetadata[index].Metadata = opts.UserDefined
+		partsMetadata[index].Fresh = true
 		partsMetadata[index].ModTime = modTime
+		partsMetadata[index].Metadata = opts.UserDefined
 	}
 
 	uploadID := mustGetUUID()

cmd/erasure-object.go

@@ -677,11 +677,7 @@ func (er erasureObjects) putMetacacheObject(ctx context.Context, key string, r *
 			onlineDisks[i] = nil
 			continue
 		}
-		if len(inlineBuffers) > 0 && inlineBuffers[i] != nil {
-			partsMetadata[i].Data = inlineBuffers[i].Bytes()
-		} else {
-			partsMetadata[i].Data = nil
-		}
+		partsMetadata[i].Data = inlineBuffers[i].Bytes()
 		partsMetadata[i].AddObjectPart(1, "", n, data.ActualSize())
 		partsMetadata[i].Erasure.AddChecksumInfo(ChecksumInfo{
 			PartNumber: 1,
@@ -695,9 +691,10 @@ func (er erasureObjects) putMetacacheObject(ctx context.Context, key string, r *
 	// Fill all the necessary metadata.
 	// Update `xl.meta` content on each disks.
 	for index := range partsMetadata {
-		partsMetadata[index].Metadata = opts.UserDefined
 		partsMetadata[index].Size = n
+		partsMetadata[index].Fresh = true
 		partsMetadata[index].ModTime = modTime
+		partsMetadata[index].Metadata = opts.UserDefined
 	}
 
 	// Set an additional header when data is inlined.
@@ -714,33 +711,11 @@ func (er erasureObjects) putMetacacheObject(ctx context.Context, key string, r *
 		}
 	}
 
-	g := errgroup.WithNErrs(len(onlineDisks))
-
-	// Rename file on all underlying storage disks.
-	for index := range onlineDisks {
-		index := index
-		g.Go(func() error {
-			if onlineDisks[index] == nil {
-				return errDiskNotFound
-			}
-			// Pick one FileInfo for a disk at index.
-			fi := partsMetadata[index]
-			// Assign index when index is initialized
-			if fi.Erasure.Index == 0 {
-				fi.Erasure.Index = index + 1
-			}
-
-			if fi.IsValid() {
-				return onlineDisks[index].WriteMetadata(ctx, minioMetaBucket, key, fi)
-			}
-			return errFileCorrupt
-		}, index)
+	if _, err = writeUniqueFileInfo(ctx, onlineDisks, minioMetaBucket, key, partsMetadata, writeQuorum); err != nil {
+		return ObjectInfo{}, toObjectErr(err, minioMetaBucket, key)
 	}
 
-	// Wait for all renames to finish.
-	errs := g.Wait()
-
-	return fi.ToObjectInfo(minioMetaBucket, key), reduceWriteQuorumErrs(ctx, errs, objectOpIgnoredErrs, writeQuorum)
+	return fi.ToObjectInfo(minioMetaBucket, key), nil
 }
 
 // PutObject - creates an object upon reading from the input stream

cmd/storage-datatypes.go

@@ -179,6 +179,8 @@ type FileInfo struct {
 
 	NumVersions      int
 	SuccessorModTime time.Time
+
+	Fresh bool // indicates this is a first time call to write FileInfo.
 }
 
 // InlineData returns true if object contents are inlined alongside its metadata.

cmd/storage-datatypes_gen.go

@@ -550,8 +550,8 @@ func (z *FileInfo) DecodeMsg(dc *msgp.Reader) (err error) {
 		err = msgp.WrapError(err)
 		return
 	}
-	if zb0001 != 24 {
-		err = msgp.ArrayError{Wanted: 24, Got: zb0001}
+	if zb0001 != 25 {
+		err = msgp.ArrayError{Wanted: 25, Got: zb0001}
 		return
 	}
 	z.Volume, err = dc.ReadString()
@@ -715,13 +715,18 @@ func (z *FileInfo) DecodeMsg(dc *msgp.Reader) (err error) {
 		err = msgp.WrapError(err, "SuccessorModTime")
 		return
 	}
+	z.Fresh, err = dc.ReadBool()
+	if err != nil {
+		err = msgp.WrapError(err, "Fresh")
+		return
+	}
 	return
 }
 
 // EncodeMsg implements msgp.Encodable
 func (z *FileInfo) EncodeMsg(en *msgp.Writer) (err error) {
-	// array header, size 24
-	err = en.Append(0xdc, 0x0, 0x18)
+	// array header, size 25
+	err = en.Append(0xdc, 0x0, 0x19)
 	if err != nil {
 		return
 	}
@@ -864,14 +869,19 @@ func (z *FileInfo) EncodeMsg(en *msgp.Writer) (err error) {
 		err = msgp.WrapError(err, "SuccessorModTime")
 		return
 	}
+	err = en.WriteBool(z.Fresh)
+	if err != nil {
+		err = msgp.WrapError(err, "Fresh")
+		return
+	}
 	return
 }
 
 // MarshalMsg implements msgp.Marshaler
 func (z *FileInfo) MarshalMsg(b []byte) (o []byte, err error) {
 	o = msgp.Require(b, z.Msgsize())
-	// array header, size 24
-	o = append(o, 0xdc, 0x0, 0x18)
+	// array header, size 25
+	o = append(o, 0xdc, 0x0, 0x19)
 	o = msgp.AppendString(o, z.Volume)
 	o = msgp.AppendString(o, z.Name)
 	o = msgp.AppendString(o, z.VersionID)
@@ -911,6 +921,7 @@ func (z *FileInfo) MarshalMsg(b []byte) (o []byte, err error) {
 	o = msgp.AppendBytes(o, z.Data)
 	o = msgp.AppendInt(o, z.NumVersions)
 	o = msgp.AppendTime(o, z.SuccessorModTime)
+	o = msgp.AppendBool(o, z.Fresh)
 	return
 }
 
@@ -922,8 +933,8 @@ func (z *FileInfo) UnmarshalMsg(bts []byte) (o []byte, err error) {
 		err = msgp.WrapError(err)
 		return
 	}
-	if zb0001 != 24 {
-		err = msgp.ArrayError{Wanted: 24, Got: zb0001}
+	if zb0001 != 25 {
+		err = msgp.ArrayError{Wanted: 25, Got: zb0001}
 		return
 	}
 	z.Volume, bts, err = msgp.ReadStringBytes(bts)
@@ -1087,6 +1098,11 @@ func (z *FileInfo) UnmarshalMsg(bts []byte) (o []byte, err error) {
 		err = msgp.WrapError(err, "SuccessorModTime")
 		return
 	}
+	z.Fresh, bts, err = msgp.ReadBoolBytes(bts)
+	if err != nil {
+		err = msgp.WrapError(err, "Fresh")
+		return
+	}
 	o = bts
 	return
 }
@@ -1104,7 +1120,7 @@ func (z *FileInfo) Msgsize() (s int) {
 	for za0003 := range z.Parts {
 		s += z.Parts[za0003].Msgsize()
 	}
-	s += z.Erasure.Msgsize() + msgp.BoolSize + msgp.StringPrefixSize + len(z.DeleteMarkerReplicationStatus) + msgp.StringPrefixSize + len(string(z.VersionPurgeStatus)) + msgp.BytesPrefixSize + len(z.Data) + msgp.IntSize + msgp.TimeSize
+	s += z.Erasure.Msgsize() + msgp.BoolSize + msgp.StringPrefixSize + len(z.DeleteMarkerReplicationStatus) + msgp.StringPrefixSize + len(string(z.VersionPurgeStatus)) + msgp.BytesPrefixSize + len(z.Data) + msgp.IntSize + msgp.TimeSize + msgp.BoolSize
 	return
 }
 

cmd/storage-rest-common.go

@@ -18,7 +18,7 @@
 package cmd
 
 const (
-	storageRESTVersion       = "v38" // Remove CheckFile API
+	storageRESTVersion       = "v39" // Add FileInfo.Fresh field
 	storageRESTVersionPrefix = SlashSeparator + storageRESTVersion
 	storageRESTPrefix        = minioReservedBucketPath + "/storage"
 )

cmd/xl-storage.go

@@ -953,6 +953,24 @@ func (s *xlStorage) UpdateMetadata(ctx context.Context, volume, path string, fi
 
 // WriteMetadata - writes FileInfo metadata for path at `xl.meta`
 func (s *xlStorage) WriteMetadata(ctx context.Context, volume, path string, fi FileInfo) error {
+	if fi.Fresh {
+		var xlMeta xlMetaV2
+		if err := xlMeta.AddVersion(fi); err != nil {
+			logger.LogIf(ctx, err)
+			return err
+		}
+		buf, err := xlMeta.AppendTo(nil)
+		if err != nil {
+			logger.LogIf(ctx, err)
+			return err
+		}
+		// First writes for special situations do not write to stable storage.
+		// this is currently used by
+		// - emphemeral objects such as objects created during listObjects() calls
+		// - newMultipartUpload() call..
+		return s.writeAll(ctx, volume, pathJoin(path, xlStorageFormatFile), buf, false)
+	}
+
 	buf, err := s.ReadAll(ctx, volume, pathJoin(path, xlStorageFormatFile))
 	if err != nil && err != errFileNotFound {
 		return err
@@ -1288,7 +1306,7 @@ func (s *xlStorage) ReadFile(ctx context.Context, volume string, path string, of
 	return int64(len(buffer)), nil
 }
 
-func (s *xlStorage) openFile(filePath string, mode int) (f *os.File, err error) {
+func (s *xlStorage) openFileSync(filePath string, mode int) (f *os.File, err error) {
 	// Create top level directories if they don't exist.
 	// with mode 0777 mkdir honors system umask.
 	if err = mkdirAll(pathutil.Dir(filePath), 0777); err != nil {
@@ -1315,6 +1333,33 @@ func (s *xlStorage) openFile(filePath string, mode int) (f *os.File, err error)
 	return w, nil
 }
 
+func (s *xlStorage) openFileNoSync(filePath string, mode int) (f *os.File, err error) {
+	// Create top level directories if they don't exist.
+	// with mode 0777 mkdir honors system umask.
+	if err = mkdirAll(pathutil.Dir(filePath), 0777); err != nil {
+		return nil, osErrToFileErr(err)
+	}
+
+	w, err := OpenFile(filePath, mode, 0666)
+	if err != nil {
+		// File path cannot be verified since one of the parents is a file.
+		switch {
+		case isSysErrIsDir(err):
+			return nil, errIsNotRegular
+		case osIsPermission(err):
+			return nil, errFileAccessDenied
+		case isSysErrIO(err):
+			return nil, errFaultyDisk
+		case isSysErrTooManyFiles(err):
+			return nil, errTooManyOpenFiles
+		default:
+			return nil, err
+		}
+	}
+
+	return w, nil
+}
+
 // To support O_DIRECT reads for erasure backends.
 type odirectReader struct {
 	f         *os.File
@@ -1525,7 +1570,7 @@ func (s *xlStorage) CreateFile(ctx context.Context, volume, path string, fileSiz
 	if fileSize >= 0 && fileSize <= smallFileThreshold {
 		// For streams smaller than 128KiB we simply write them as O_DSYNC (fdatasync)
 		// and not O_DIRECT to avoid the complexities of aligned I/O.
-		w, err := s.openFile(filePath, os.O_CREATE|os.O_WRONLY|os.O_EXCL)
+		w, err := s.openFileSync(filePath, os.O_CREATE|os.O_WRONLY|os.O_EXCL)
 		if err != nil {
 			return err
 		}
@@ -1585,7 +1630,7 @@ func (s *xlStorage) CreateFile(ctx context.Context, volume, path string, fileSiz
 	return nil
 }
 
-func (s *xlStorage) WriteAll(ctx context.Context, volume string, path string, b []byte) (err error) {
+func (s *xlStorage) writeAll(ctx context.Context, volume string, path string, b []byte, sync bool) (err error) {
 	volumeDir, err := s.getVolDir(volume)
 	if err != nil {
 		return err
@@ -1596,7 +1641,12 @@ func (s *xlStorage) WriteAll(ctx context.Context, volume string, path string, b
 		return err
 	}
 
-	w, err := s.openFile(filePath, os.O_CREATE|os.O_WRONLY|os.O_TRUNC)
+	var w *os.File
+	if sync {
+		w, err = s.openFileSync(filePath, os.O_CREATE|os.O_WRONLY|os.O_TRUNC)
+	} else {
+		w, err = s.openFileNoSync(filePath, os.O_CREATE|os.O_WRONLY|os.O_TRUNC)
+	}
 	if err != nil {
 		return err
 	}
@@ -1614,6 +1664,10 @@ func (s *xlStorage) WriteAll(ctx context.Context, volume string, path string, b
 	return nil
 }
 
+func (s *xlStorage) WriteAll(ctx context.Context, volume string, path string, b []byte) (err error) {
+	return s.writeAll(ctx, volume, path, b, true)
+}
+
 // AppendFile - append a byte array at path, if file doesn't exist at
 // path this call explicitly creates it.
 func (s *xlStorage) AppendFile(ctx context.Context, volume string, path string, buf []byte) (err error) {
@@ -1642,7 +1696,7 @@ func (s *xlStorage) AppendFile(ctx context.Context, volume string, path string,
 	var w *os.File
 	// Create file if not found. Not doing O_DIRECT here to avoid the code that does buffer aligned writes.
 	// AppendFile() is only used by healing code to heal objects written in old format.
-	w, err = s.openFile(filePath, os.O_CREATE|os.O_APPEND|os.O_WRONLY)
+	w, err = s.openFileSync(filePath, os.O_CREATE|os.O_APPEND|os.O_WRONLY)
 	if err != nil {
 		return err
 	}