Merge remote-tracking branch 'upstream/master'

5 files changed, 442 insertions, 0 deletions

diff --git a/weed/util/chunk_cache/chunk_cache.go b/weed/util/chunk_cache/chunk_cache.go
new file mode 100644
index 000000000..e1d4b639f
--- /dev/null
+++ b/weed/util/chunk_cache/chunk_cache.go
@@ -0,0 +1,113 @@
+package chunk_cache
+
+import (
+	"sync"
+
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/storage/needle"
+)
+
+const (
+	memCacheSizeLimit     = 1024 * 1024
+	onDiskCacheSizeLimit0 = memCacheSizeLimit
+	onDiskCacheSizeLimit1 = 4 * memCacheSizeLimit
+)
+
+// a global cache for recently accessed file chunks
+type ChunkCache struct {
+	memCache   *ChunkCacheInMemory
+	diskCaches []*OnDiskCacheLayer
+	sync.RWMutex
+}
+
+func NewChunkCache(maxEntries int64, dir string, diskSizeMB int64) *ChunkCache {
+
+	c := &ChunkCache{
+		memCache: NewChunkCacheInMemory(maxEntries),
+	}
+	c.diskCaches = make([]*OnDiskCacheLayer, 3)
+	c.diskCaches[0] = NewOnDiskCacheLayer(dir, "c0_1", diskSizeMB/4, 4)
+	c.diskCaches[1] = NewOnDiskCacheLayer(dir, "c1_4", diskSizeMB/4, 4)
+	c.diskCaches[2] = NewOnDiskCacheLayer(dir, "cache", diskSizeMB/2, 4)
+
+	return c
+}
+
+func (c *ChunkCache) GetChunk(fileId string, chunkSize uint64) (data []byte) {
+	if c == nil {
+		return
+	}
+
+	c.RLock()
+	defer c.RUnlock()
+
+	return c.doGetChunk(fileId, chunkSize)
+}
+
+func (c *ChunkCache) doGetChunk(fileId string, chunkSize uint64) (data []byte) {
+
+	if chunkSize < memCacheSizeLimit {
+		if data = c.memCache.GetChunk(fileId); data != nil {
+			return data
+		}
+	}
+
+	fid, err := needle.ParseFileIdFromString(fileId)
+	if err != nil {
+		glog.Errorf("failed to parse file id %s", fileId)
+		return nil
+	}
+
+	for _, diskCache := range c.diskCaches {
+		data := diskCache.getChunk(fid.Key)
+		if len(data) != 0 {
+			return data
+		}
+	}
+
+	return nil
+
+}
+
+func (c *ChunkCache) SetChunk(fileId string, data []byte) {
+	if c == nil {
+		return
+	}
+	c.Lock()
+	defer c.Unlock()
+
+	c.doSetChunk(fileId, data)
+}
+
+func (c *ChunkCache) doSetChunk(fileId string, data []byte) {
+
+	if len(data) < memCacheSizeLimit {
+		c.memCache.SetChunk(fileId, data)
+	}
+
+	fid, err := needle.ParseFileIdFromString(fileId)
+	if err != nil {
+		glog.Errorf("failed to parse file id %s", fileId)
+		return
+	}
+
+	if len(data) < onDiskCacheSizeLimit0 {
+		c.diskCaches[0].setChunk(fid.Key, data)
+	} else if len(data) < onDiskCacheSizeLimit1 {
+		c.diskCaches[1].setChunk(fid.Key, data)
+	} else {
+		c.diskCaches[2].setChunk(fid.Key, data)
+	}
+
+}
+
+func (c *ChunkCache) Shutdown() {
+	if c == nil {
+		return
+	}
+	c.Lock()
+	defer c.Unlock()
+	for _, diskCache := range c.diskCaches {
+		diskCache.shutdown()
+	}
+}
diff --git a/weed/util/chunk_cache/chunk_cache_in_memory.go b/weed/util/chunk_cache/chunk_cache_in_memory.go
new file mode 100644
index 000000000..931e45e9a
--- /dev/null
+++ b/weed/util/chunk_cache/chunk_cache_in_memory.go
@@ -0,0 +1,36 @@
+package chunk_cache
+
+import (
+	"time"
+
+	"github.com/karlseguin/ccache"
+)
+
+// a global cache for recently accessed file chunks
+type ChunkCacheInMemory struct {
+	cache *ccache.Cache
+}
+
+func NewChunkCacheInMemory(maxEntries int64) *ChunkCacheInMemory {
+	pruneCount := maxEntries >> 3
+	if pruneCount <= 0 {
+		pruneCount = 500
+	}
+	return &ChunkCacheInMemory{
+		cache: ccache.New(ccache.Configure().MaxSize(maxEntries).ItemsToPrune(uint32(pruneCount))),
+	}
+}
+
+func (c *ChunkCacheInMemory) GetChunk(fileId string) []byte {
+	item := c.cache.Get(fileId)
+	if item == nil {
+		return nil
+	}
+	data := item.Value().([]byte)
+	item.Extend(time.Hour)
+	return data
+}
+
+func (c *ChunkCacheInMemory) SetChunk(fileId string, data []byte) {
+	c.cache.Set(fileId, data, time.Hour)
+}
diff --git a/weed/util/chunk_cache/chunk_cache_on_disk.go b/weed/util/chunk_cache/chunk_cache_on_disk.go
new file mode 100644
index 000000000..2c7ef8d39
--- /dev/null
+++ b/weed/util/chunk_cache/chunk_cache_on_disk.go
@@ -0,0 +1,145 @@
+package chunk_cache
+
+import (
+	"fmt"
+	"os"
+	"time"
+
+	"github.com/syndtr/goleveldb/leveldb/opt"
+
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/storage"
+	"github.com/chrislusf/seaweedfs/weed/storage/backend"
+	"github.com/chrislusf/seaweedfs/weed/storage/types"
+	"github.com/chrislusf/seaweedfs/weed/util"
+)
+
+// This implements an on disk cache
+// The entries are an FIFO with a size limit
+
+type ChunkCacheVolume struct {
+	DataBackend backend.BackendStorageFile
+	nm          storage.NeedleMapper
+	fileName    string
+	smallBuffer []byte
+	sizeLimit   int64
+	lastModTime time.Time
+	fileSize    int64
+}
+
+func LoadOrCreateChunkCacheVolume(fileName string, preallocate int64) (*ChunkCacheVolume, error) {
+
+	v := &ChunkCacheVolume{
+		smallBuffer: make([]byte, types.NeedlePaddingSize),
+		fileName:    fileName,
+		sizeLimit:   preallocate,
+	}
+
+	var err error
+
+	if exists, canRead, canWrite, modTime, fileSize := util.CheckFile(v.fileName + ".dat"); exists {
+		if !canRead {
+			return nil, fmt.Errorf("cannot read cache file %s.dat", v.fileName)
+		}
+		if !canWrite {
+			return nil, fmt.Errorf("cannot write cache file %s.dat", v.fileName)
+		}
+		if dataFile, err := os.OpenFile(v.fileName+".dat", os.O_RDWR|os.O_CREATE, 0644); err != nil {
+			return nil, fmt.Errorf("cannot create cache file %s.dat: %v", v.fileName, err)
+		} else {
+			v.DataBackend = backend.NewDiskFile(dataFile)
+			v.lastModTime = modTime
+			v.fileSize = fileSize
+		}
+	} else {
+		if v.DataBackend, err = backend.CreateVolumeFile(v.fileName+".dat", preallocate, 0); err != nil {
+			return nil, fmt.Errorf("cannot create cache file %s.dat: %v", v.fileName, err)
+		}
+		v.lastModTime = time.Now()
+	}
+
+	var indexFile *os.File
+	if indexFile, err = os.OpenFile(v.fileName+".idx", os.O_RDWR|os.O_CREATE, 0644); err != nil {
+		return nil, fmt.Errorf("cannot write cache index %s.idx: %v", v.fileName, err)
+	}
+
+	glog.V(0).Infoln("loading leveldb", v.fileName+".ldb")
+	opts := &opt.Options{
+		BlockCacheCapacity:            2 * 1024 * 1024, // default value is 8MiB
+		WriteBuffer:                   1 * 1024 * 1024, // default value is 4MiB
+		CompactionTableSizeMultiplier: 10,              // default value is 1
+	}
+	if v.nm, err = storage.NewLevelDbNeedleMap(v.fileName+".ldb", indexFile, opts); err != nil {
+		return nil, fmt.Errorf("loading leveldb %s error: %v", v.fileName+".ldb", err)
+	}
+
+	return v, nil
+
+}
+
+func (v *ChunkCacheVolume) Shutdown() {
+	if v.DataBackend != nil {
+		v.DataBackend.Close()
+		v.DataBackend = nil
+	}
+	if v.nm != nil {
+		v.nm.Close()
+		v.nm = nil
+	}
+}
+
+func (v *ChunkCacheVolume) destroy() {
+	v.Shutdown()
+	os.Remove(v.fileName + ".dat")
+	os.Remove(v.fileName + ".idx")
+	os.RemoveAll(v.fileName + ".ldb")
+}
+
+func (v *ChunkCacheVolume) Reset() (*ChunkCacheVolume, error) {
+	v.destroy()
+	return LoadOrCreateChunkCacheVolume(v.fileName, v.sizeLimit)
+}
+
+func (v *ChunkCacheVolume) GetNeedle(key types.NeedleId) ([]byte, error) {
+
+	nv, ok := v.nm.Get(key)
+	if !ok {
+		return nil, storage.ErrorNotFound
+	}
+	data := make([]byte, nv.Size)
+	if readSize, readErr := v.DataBackend.ReadAt(data, nv.Offset.ToAcutalOffset()); readErr != nil {
+		return nil, fmt.Errorf("read %s.dat [%d,%d): %v",
+			v.fileName, nv.Offset.ToAcutalOffset(), nv.Offset.ToAcutalOffset()+int64(nv.Size), readErr)
+	} else {
+		if readSize != int(nv.Size) {
+			return nil, fmt.Errorf("read %d, expected %d", readSize, nv.Size)
+		}
+	}
+
+	return data, nil
+}
+
+func (v *ChunkCacheVolume) WriteNeedle(key types.NeedleId, data []byte) error {
+
+	offset := v.fileSize
+
+	written, err := v.DataBackend.WriteAt(data, offset)
+	if err != nil {
+		return err
+	} else if written != len(data) {
+		return fmt.Errorf("partial written %d, expected %d", written, len(data))
+	}
+
+	v.fileSize += int64(written)
+	extraSize := written % types.NeedlePaddingSize
+	if extraSize != 0 {
+		v.DataBackend.WriteAt(v.smallBuffer[:types.NeedlePaddingSize-extraSize], offset+int64(written))
+		v.fileSize += int64(types.NeedlePaddingSize - extraSize)
+	}
+
+	if err := v.nm.Put(key, types.ToOffset(offset), uint32(len(data))); err != nil {
+		glog.V(4).Infof("failed to save in needle map %d: %v", key, err)
+	}
+
+	return nil
+}
diff --git a/weed/util/chunk_cache/chunk_cache_on_disk_test.go b/weed/util/chunk_cache/chunk_cache_on_disk_test.go
new file mode 100644
index 000000000..f061f2ba2
--- /dev/null
+++ b/weed/util/chunk_cache/chunk_cache_on_disk_test.go
@@ -0,0 +1,59 @@
+package chunk_cache
+
+import (
+	"bytes"
+	"fmt"
+	"io/ioutil"
+	"math/rand"
+	"os"
+	"testing"
+)
+
+func TestOnDisk(t *testing.T) {
+
+	tmpDir, _ := ioutil.TempDir("", "c")
+	defer os.RemoveAll(tmpDir)
+
+	totalDiskSizeMb := int64(32)
+
+	cache := NewChunkCache(0, tmpDir, totalDiskSizeMb)
+
+	writeCount := 5
+	type test_data struct {
+		data   []byte
+		fileId string
+		size   uint64
+	}
+	testData := make([]*test_data, writeCount)
+	for i := 0; i < writeCount; i++ {
+		buff := make([]byte, 1024*1024)
+		rand.Read(buff)
+		testData[i] = &test_data{
+			data:   buff,
+			fileId: fmt.Sprintf("1,%daabbccdd", i+1),
+			size:   uint64(len(buff)),
+		}
+		cache.SetChunk(testData[i].fileId, testData[i].data)
+	}
+
+	for i := 0; i < writeCount; i++ {
+		data := cache.GetChunk(testData[i].fileId, testData[i].size)
+		if bytes.Compare(data, testData[i].data) != 0 {
+			t.Errorf("failed to write to and read from cache: %d", i)
+		}
+	}
+
+	cache.Shutdown()
+
+	cache = NewChunkCache(0, tmpDir, totalDiskSizeMb)
+
+	for i := 0; i < writeCount; i++ {
+		data := cache.GetChunk(testData[i].fileId, testData[i].size)
+		if bytes.Compare(data, testData[i].data) != 0 {
+			t.Errorf("failed to write to and read from cache: %d", i)
+		}
+	}
+
+	cache.Shutdown()
+
+}
diff --git a/weed/util/chunk_cache/on_disk_cache_layer.go b/weed/util/chunk_cache/on_disk_cache_layer.go
new file mode 100644
index 000000000..9cf8e3ab2
--- /dev/null
+++ b/weed/util/chunk_cache/on_disk_cache_layer.go
@@ -0,0 +1,89 @@
+package chunk_cache
+
+import (
+	"fmt"
+	"path"
+	"sort"
+
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/storage"
+	"github.com/chrislusf/seaweedfs/weed/storage/types"
+)
+
+type OnDiskCacheLayer struct {
+	diskCaches []*ChunkCacheVolume
+}
+
+func NewOnDiskCacheLayer(dir, namePrefix string, diskSizeMB int64, segmentCount int) *OnDiskCacheLayer {
+
+	volumeCount, volumeSize := int(diskSizeMB/30000), int64(30000)
+	if volumeCount < segmentCount {
+		volumeCount, volumeSize = segmentCount, diskSizeMB/int64(segmentCount)
+	}
+
+	c := &OnDiskCacheLayer{}
+	for i := 0; i < volumeCount; i++ {
+		fileName := path.Join(dir, fmt.Sprintf("%s_%d", namePrefix, i))
+		diskCache, err := LoadOrCreateChunkCacheVolume(fileName, volumeSize*1024*1024)
+		if err != nil {
+			glog.Errorf("failed to add cache %s : %v", fileName, err)
+		} else {
+			c.diskCaches = append(c.diskCaches, diskCache)
+		}
+	}
+
+	// keep newest cache to the front
+	sort.Slice(c.diskCaches, func(i, j int) bool {
+		return c.diskCaches[i].lastModTime.After(c.diskCaches[j].lastModTime)
+	})
+
+	return c
+}
+
+func (c *OnDiskCacheLayer) setChunk(needleId types.NeedleId, data []byte) {
+
+	if c.diskCaches[0].fileSize+int64(len(data)) > c.diskCaches[0].sizeLimit {
+		t, resetErr := c.diskCaches[len(c.diskCaches)-1].Reset()
+		if resetErr != nil {
+			glog.Errorf("failed to reset cache file %s", c.diskCaches[len(c.diskCaches)-1].fileName)
+			return
+		}
+		for i := len(c.diskCaches) - 1; i > 0; i-- {
+			c.diskCaches[i] = c.diskCaches[i-1]
+		}
+		c.diskCaches[0] = t
+	}
+
+	c.diskCaches[0].WriteNeedle(needleId, data)
+
+}
+
+func (c *OnDiskCacheLayer) getChunk(needleId types.NeedleId) (data []byte) {
+
+	var err error
+
+	for _, diskCache := range c.diskCaches {
+		data, err = diskCache.GetNeedle(needleId)
+		if err == storage.ErrorNotFound {
+			continue
+		}
+		if err != nil {
+			glog.Errorf("failed to read cache file %s id %d", diskCache.fileName, needleId)
+			continue
+		}
+		if len(data) != 0 {
+			return
+		}
+	}
+
+	return nil
+
+}
+
+func (c *OnDiskCacheLayer) shutdown() {
+
+	for _, diskCache := range c.diskCaches {
+		diskCache.Shutdown()
+	}
+
+}