Feat:merge small chunk (#6049)

* fix:mount deadlock

* feat: merge small chunk

* adjust MergeChunkMinCount

* fix

---------

Co-authored-by: zemul <zhouzemiao@ihuman.com>

weed/server/filer_server_handlers_write_autochunk.go

@@ -99,7 +99,7 @@ func (fs *FilerServer) doPostAutoChunk(ctx context.Context, w http.ResponseWrite
 		return
 	}
 
-	fileChunks, md5Hash, chunkOffset, err, smallContent := fs.uploadReaderToChunks(w, r, part1, chunkSize, fileName, contentType, contentLength, so)
+	fileChunks, md5Hash, chunkOffset, err, smallContent := fs.uploadRequestToChunks(w, r, part1, chunkSize, fileName, contentType, contentLength, so)
 	if err != nil {
 		return nil, nil, err
 	}
@@ -130,7 +130,8 @@ func (fs *FilerServer) doPutAutoChunk(ctx context.Context, w http.ResponseWriter
 		return nil, nil, err
 	}
 
-	fileChunks, md5Hash, chunkOffset, err, smallContent := fs.uploadReaderToChunks(w, r, r.Body, chunkSize, fileName, contentType, contentLength, so)
+	fileChunks, md5Hash, chunkOffset, err, smallContent := fs.uploadRequestToChunks(w, r, r.Body, chunkSize, fileName, contentType, contentLength, so)
+
 	if err != nil {
 		return nil, nil, err
 	}

weed/server/filer_server_handlers_write_merge.go

@@ -1,11 +1,65 @@
 package weed_server
 
 import (
+	"github.com/seaweedfs/seaweedfs/weed/filer"
+	"github.com/seaweedfs/seaweedfs/weed/glog"
 	"github.com/seaweedfs/seaweedfs/weed/operation"
 	"github.com/seaweedfs/seaweedfs/weed/pb/filer_pb"
+	"github.com/seaweedfs/seaweedfs/weed/stats"
+	"io"
+	"math"
 )
 
+const MergeChunkMinCount int = 1000
+
 func (fs *FilerServer) maybeMergeChunks(so *operation.StorageOption, inputChunks []*filer_pb.FileChunk) (mergedChunks []*filer_pb.FileChunk, err error) {
-	//TODO merge consecutive smaller chunks into a large chunk to reduce number of chunks
+	// Only merge small chunks more than half of the file
+	var chunkSize = fs.option.MaxMB * 1024 * 1024
+	var smallChunk, sumChunk int
+	var minOffset int64 = math.MaxInt64
+	for _, chunk := range inputChunks {
+		if chunk.IsChunkManifest {
+			continue
+		}
+		if chunk.Size < uint64(chunkSize/2) {
+			smallChunk++
+			if chunk.Offset < minOffset {
+				minOffset = chunk.Offset
+			}
+		}
+		sumChunk++
+	}
+	if smallChunk < MergeChunkMinCount || smallChunk < sumChunk/2 {
 		return inputChunks, nil
 	}
+
+	return fs.mergeChunks(so, inputChunks, minOffset)
+}
+
+func (fs *FilerServer) mergeChunks(so *operation.StorageOption, inputChunks []*filer_pb.FileChunk, chunkOffset int64) (mergedChunks []*filer_pb.FileChunk, mergeErr error) {
+	chunkedFileReader := filer.NewChunkStreamReaderFromFiler(fs.filer.MasterClient, inputChunks)
+	_, mergeErr = chunkedFileReader.Seek(chunkOffset, io.SeekCurrent)
+	if mergeErr != nil {
+		return nil, mergeErr
+	}
+	mergedChunks, _, _, mergeErr, _ = fs.uploadReaderToChunks(chunkedFileReader, chunkOffset, int32(fs.option.MaxMB*1024*1024), "", "", true, so)
+	if mergeErr != nil {
+		return
+	}
+
+	stats.FilerHandlerCounter.WithLabelValues(stats.ChunkMerge).Inc()
+	for _, chunk := range inputChunks {
+		if chunk.Offset < chunkOffset || chunk.IsChunkManifest {
+			mergedChunks = append(mergedChunks, chunk)
+		}
+	}
+
+	garbage, err := filer.MinusChunks(fs.lookupFileId, inputChunks, mergedChunks)
+	if err != nil {
+		glog.Errorf("Failed to resolve old entry chunks when delete old entry chunks. new: %s, old: %s",
+			mergedChunks, inputChunks)
+		return
+	}
+	fs.filer.DeleteChunksNotRecursive(garbage)
+	return
+}

weed/server/filer_server_handlers_write_upload.go

@@ -27,7 +27,7 @@ var bufPool = sync.Pool{
 	},
 }
 
-func (fs *FilerServer) uploadReaderToChunks(w http.ResponseWriter, r *http.Request, reader io.Reader, chunkSize int32, fileName, contentType string, contentLength int64, so *operation.StorageOption) (fileChunks []*filer_pb.FileChunk, md5Hash hash.Hash, chunkOffset int64, uploadErr error, smallContent []byte) {
+func (fs *FilerServer) uploadRequestToChunks(w http.ResponseWriter, r *http.Request, reader io.Reader, chunkSize int32, fileName, contentType string, contentLength int64, so *operation.StorageOption) (fileChunks []*filer_pb.FileChunk, md5Hash hash.Hash, chunkOffset int64, uploadErr error, smallContent []byte) {
 	query := r.URL.Query()
 
 	isAppend := isAppend(r)
@@ -45,7 +45,13 @@ func (fs *FilerServer) uploadReaderToChunks(w http.ResponseWriter, r *http.Reque
 		chunkOffset = offsetInt
 	}
 
+	return fs.uploadReaderToChunks(reader, chunkOffset, chunkSize, fileName, contentType, isAppend, so)
+}
+
+func (fs *FilerServer) uploadReaderToChunks(reader io.Reader, startOffset int64, chunkSize int32, fileName, contentType string, isAppend bool, so *operation.StorageOption) (fileChunks []*filer_pb.FileChunk, md5Hash hash.Hash, chunkOffset int64, uploadErr error, smallContent []byte) {
+
 	md5Hash = md5.New()
+	chunkOffset = startOffset
 	var partReader = io.NopCloser(io.TeeReader(reader, md5Hash))
 
 	var wg sync.WaitGroup

weed/stats/metrics_names.go

@@ -33,6 +33,8 @@ const (
 	ChunkProxy         = "chunkProxy"
 	ChunkAssign        = "chunkAssign"
 	ChunkUpload        = "chunkUpload"
+	ChunkMerge         = "chunkMerge"
+
 	ChunkDoUploadRetry       = "chunkDoUploadRetry"
 	ChunkUploadRetry         = "chunkUploadRetry"
 	ChunkAssignRetry         = "chunkAssignRetry"