filer: parallel data upload

1 files changed, 34 insertions, 19 deletions

diff --git a/weed/server/filer_server_handlers_write_upload.go b/weed/server/filer_server_handlers_write_upload.go
index d093676bc..f2b996ec5 100644
--- a/weed/server/filer_server_handlers_write_upload.go
+++ b/weed/server/filer_server_handlers_write_upload.go
@@ -7,6 +7,7 @@ import (
 	"io"
 	"io/ioutil"
 	"net/http"
+	"sort"
 	"strings"
 	"sync"
 	"time"
@@ -34,10 +35,15 @@ func (fs *FilerServer) uploadReaderToChunks(w http.ResponseWriter, r *http.Reque
 
 	chunkOffset := int64(0)
 	var smallContent []byte
+	var uploadErr error
 
-	bytesBuffer := bufPool.Get().(*bytes.Buffer)
-	defer bufPool.Put(bytesBuffer)
+	var wg sync.WaitGroup
 	for {
+
+		// need to throttle this for large files
+		bytesBuffer := bufPool.Get().(*bytes.Buffer)
+		defer bufPool.Put(bytesBuffer)
+
 		limitedReader := io.LimitReader(partReader, int64(chunkSize))
 
 		bytesBuffer.Reset()
@@ -45,8 +51,8 @@ func (fs *FilerServer) uploadReaderToChunks(w http.ResponseWriter, r *http.Reque
 		dataSize, err := bytesBuffer.ReadFrom(limitedReader)
 
 		// data, err := ioutil.ReadAll(limitedReader)
-		if err != nil {
-			return nil, nil, 0, err, nil
+		if err != nil || dataSize == 0 {
+			return nil, md5Hash, 0, err, nil
 		}
 		if chunkOffset == 0 && !isAppend(r) {
 			if dataSize < fs.option.SaveToFilerLimit || strings.HasPrefix(r.URL.Path, filer.DirectoryEtcRoot) && dataSize < 4*1024 {
@@ -57,30 +63,39 @@ func (fs *FilerServer) uploadReaderToChunks(w http.ResponseWriter, r *http.Reque
 			}
 		}
 
-		chunk, uploadErr := fs.dataToChunk(fileName, contentType, bytesBuffer.Bytes(), chunkOffset, so, md5Hash)
-
-		if uploadErr != nil {
-			return nil, nil, 0, uploadErr, nil
-		}
-
-		// if last chunk exhausted the reader exactly at the border
-		if chunk == nil {
-			break
-		}
-		// Save to chunk manifest structure
-		fileChunks = append(fileChunks, chunk)
+		wg.Add(1)
+		go func(offset int64) {
+			defer wg.Done()
 
-		glog.V(4).Infof("uploaded %s chunk %d to %s [%d,%d)", fileName, len(fileChunks), chunk.FileId, chunkOffset, chunkOffset+int64(chunk.Size))
+			chunk, toChunkErr := fs.dataToChunk(fileName, contentType, bytesBuffer.Bytes(), offset, so, md5Hash)
+			if toChunkErr != nil {
+				uploadErr = toChunkErr
+			}
+			if chunk != nil {
+				fileChunks = append(fileChunks, chunk)
+				glog.V(4).Infof("uploaded %s chunk %d to %s [%d,%d)", fileName, len(fileChunks), chunk.FileId, offset, offset+int64(chunk.Size))
+			}
+		}(chunkOffset)
 
 		// reset variables for the next chunk
-		chunkOffset = chunkOffset + int64(chunk.Size)
+		chunkOffset = chunkOffset + dataSize
 
 		// if last chunk was not at full chunk size, but already exhausted the reader
-		if int64(chunk.Size) < int64(chunkSize) {
+		if dataSize < int64(chunkSize) {
 			break
 		}
 	}
 
+	wg.Wait()
+
+	if uploadErr != nil {
+		return nil, md5Hash, 0, uploadErr, nil
+	}
+
+	sort.Slice(fileChunks, func(i, j int) bool {
+		return fileChunks[i].Offset < fileChunks[j].Offset
+	})
+
 	return fileChunks, md5Hash, chunkOffset, nil, smallContent
 }