X-Git-Url: https://git.arvados.org/arvados.git/blobdiff_plain/9e5b98e8f5f4727856b53447191f9c06e3da2ba6..aa798e6915bd39296bca8f62fd3476293c25f436:/services/crunch-run/upload.go

diff --git a/services/crunch-run/upload.go b/services/crunch-run/upload.go
index a068a2a77b..b54e336c2d 100644
--- a/services/crunch-run/upload.go
+++ b/services/crunch-run/upload.go
@@ -1,3 +1,7 @@
+// Copyright (C) The Arvados Authors. All rights reserved.
+//
+// SPDX-License-Identifier: AGPL-3.0
+
 package main
 
 // Originally based on sdk/go/crunchrunner/upload.go
@@ -14,14 +18,15 @@ import (
 	"crypto/md5"
 	"errors"
 	"fmt"
-	"git.curoverse.com/arvados.git/sdk/go/keepclient"
-	"git.curoverse.com/arvados.git/sdk/go/manifest"
 	"io"
 	"log"
 	"os"
 	"path/filepath"
 	"strings"
 	"sync"
+
+	"git.curoverse.com/arvados.git/sdk/go/keepclient"
+	"git.curoverse.com/arvados.git/sdk/go/manifest"
 )
 
 // Block is a data block in a manifest stream
@@ -87,27 +92,50 @@ func (m *CollectionFileWriter) NewFile(fn string) {
 	m.fn = fn
 }
 
-func (m *CollectionFileWriter) goUpload() {
+func (m *CollectionFileWriter) goUpload(workers chan struct{}) {
+	var mtx sync.Mutex
+	var wg sync.WaitGroup
+
 	var errors []error
 	uploader := m.uploader
 	finish := m.finish
 	for block := range uploader {
-		hash := fmt.Sprintf("%x", md5.Sum(block.data[0:block.offset]))
-		signedHash, _, err := m.IKeepClient.PutHB(hash, block.data[0:block.offset])
-		if err != nil {
-			errors = append(errors, err)
-		} else {
-			m.ManifestStream.Blocks = append(m.ManifestStream.Blocks, signedHash)
-		}
+		mtx.Lock()
+		m.ManifestStream.Blocks = append(m.ManifestStream.Blocks, "")
+		blockIndex := len(m.ManifestStream.Blocks) - 1
+		mtx.Unlock()
+
+		workers <- struct{}{} // wait for an available worker slot
+		wg.Add(1)
+
+		go func(block *Block, blockIndex int) {
+			hash := fmt.Sprintf("%x", md5.Sum(block.data[0:block.offset]))
+			signedHash, _, err := m.IKeepClient.PutHB(hash, block.data[0:block.offset])
+			<-workers
+
+			mtx.Lock()
+			if err != nil {
+				errors = append(errors, err)
+			} else {
+				m.ManifestStream.Blocks[blockIndex] = signedHash
+			}
+			mtx.Unlock()
+
+			wg.Done()
+		}(block, blockIndex)
 	}
+	wg.Wait()
+
 	finish <- errors
 }
 
 // CollectionWriter implements creating new Keep collections by opening files
 // and writing to them.
 type CollectionWriter struct {
+	MaxWriters int
 	IKeepClient
 	Streams []*CollectionFileWriter
+	workers chan struct{}
 	mtx     sync.Mutex
 }
 
@@ -134,10 +162,18 @@ func (m *CollectionWriter) Open(path string) io.WriteCloser {
 		make(chan *Block),
 		make(chan []error),
 		fn}
-	go fw.goUpload()
 
 	m.mtx.Lock()
 	defer m.mtx.Unlock()
+	if m.workers == nil {
+		if m.MaxWriters < 1 {
+			m.MaxWriters = 2
+		}
+		m.workers = make(chan struct{}, m.MaxWriters)
+	}
+
+	go fw.goUpload(m.workers)
+
 	m.Streams = append(m.Streams, fw)
 
 	return fw
@@ -218,29 +254,36 @@ func (m *CollectionWriter) ManifestText() (mt string, err error) {
 }
 
 type WalkUpload struct {
+	MaxWriters  int
 	kc          IKeepClient
 	stripPrefix string
 	streamMap   map[string]*CollectionFileWriter
 	status      *log.Logger
+	workers     chan struct{}
+	mtx         sync.Mutex
 }
 
-// WalkFunc walks a directory tree, uploads each file found and adds it to the
-// CollectionWriter.
-func (m *WalkUpload) WalkFunc(path string, info os.FileInfo, err error) error {
-
-	if info.IsDir() {
-		return nil
-	}
-
+func (m *WalkUpload) UploadFile(path string, sourcePath string) error {
 	var dir string
-	if len(path) > (len(m.stripPrefix) + len(info.Name()) + 1) {
-		dir = path[len(m.stripPrefix)+1 : (len(path) - len(info.Name()) - 1)]
+	basename := filepath.Base(path)
+	if len(path) > (len(m.stripPrefix) + len(basename) + 1) {
+		dir = path[len(m.stripPrefix)+1 : (len(path) - len(basename) - 1)]
 	}
 	if dir == "" {
 		dir = "."
 	}
 
-	fn := path[(len(path) - len(info.Name())):]
+	fn := path[(len(path) - len(basename)):]
+
+	info, err := os.Stat(sourcePath)
+	if err != nil {
+		return err
+	}
+	file, err := os.Open(sourcePath)
+	if err != nil {
+		return err
+	}
+	defer file.Close()
 
 	if m.streamMap[dir] == nil {
 		m.streamMap[dir] = &CollectionFileWriter{
@@ -252,7 +295,17 @@ func (m *WalkUpload) WalkFunc(path string, info os.FileInfo, err error) error {
 			make(chan *Block),
 			make(chan []error),
 			""}
-		go m.streamMap[dir].goUpload()
+
+		m.mtx.Lock()
+		if m.workers == nil {
+			if m.MaxWriters < 1 {
+				m.MaxWriters = 2
+			}
+			m.workers = make(chan struct{}, m.MaxWriters)
+		}
+		m.mtx.Unlock()
+
+		go m.streamMap[dir].goUpload(m.workers)
 	}
 
 	fileWriter := m.streamMap[dir]
@@ -260,16 +313,11 @@ func (m *WalkUpload) WalkFunc(path string, info os.FileInfo, err error) error {
 	// Reset the CollectionFileWriter for a new file
 	fileWriter.NewFile(fn)
 
-	file, err := os.Open(path)
-	if err != nil {
-		return err
-	}
-	defer file.Close()
-
 	m.status.Printf("Uploading %v/%v (%v bytes)", dir, fn, info.Size())
 
 	_, err = io.Copy(fileWriter, file)
 	if err != nil {
+		m.status.Printf("Uh oh")
 		return err
 	}
 
@@ -279,20 +327,15 @@ func (m *WalkUpload) WalkFunc(path string, info os.FileInfo, err error) error {
 	return nil
 }
 
-func (cw *CollectionWriter) WriteTree(root string, status *log.Logger) (manifest string, err error) {
+func (cw *CollectionWriter) BeginUpload(root string, status *log.Logger) *WalkUpload {
 	streamMap := make(map[string]*CollectionFileWriter)
-	wu := &WalkUpload{cw.IKeepClient, root, streamMap, status}
-	err = filepath.Walk(root, wu.WalkFunc)
-
-	if err != nil {
-		return "", err
-	}
+	return &WalkUpload{0, cw.IKeepClient, root, streamMap, status, nil, sync.Mutex{}}
+}
 
+func (cw *CollectionWriter) EndUpload(wu *WalkUpload) {
 	cw.mtx.Lock()
-	for _, st := range streamMap {
+	for _, st := range wu.streamMap {
 		cw.Streams = append(cw.Streams, st)
 	}
 	cw.mtx.Unlock()
-
-	return cw.ManifestText()
 }