seaweedfs/go/storage/volume.go
chrislusf 1b6ab2f6af Add boltdb for volume needle map
boltdb is fairly slow to write, about 6 minutes for recreating index
for 1553934 files. Boltdb loads 1,553,934 x 16 = 24,862,944bytes from
disk, and generate the boltdb as large as 134,217,728 bytes in 6
minutes.

To compare, for leveldb, it recreates index in leveldb as large as
27,188,148 bytes in 8 seconds.
For in memory version, it loads the index in

To test the memory consumption, the leveldb or boltdb index are
created. And the server is restarted. Using the benchmark tool to read
lots of files. There are 7 volumes in benchmark collection, each with
about 1553K files.
For leveldb, the memory starts at 142,884KB, and stays at 179,340KB.
For boltdb, the memory starts at 73,756KB, and stays at 144,564KB.
For in-memory, the memory starts at 368,152KB, and stays at 448,032KB.
2015-03-29 11:04:32 -07:00

393 lines
11 KiB
Go

package storage
import (
"bytes"
"errors"
"fmt"
"io"
"os"
"path"
"sync"
"time"
"github.com/chrislusf/weed-fs/go/glog"
)
type Volume struct {
Id VolumeId
dir string
Collection string
dataFile *os.File
nm NeedleMapper
needleMapKind NeedleMapType
readOnly bool
SuperBlock
accessLock sync.Mutex
lastModifiedTime uint64 //unix time in seconds
}
func NewVolume(dirname string, collection string, id VolumeId, needleMapKind NeedleMapType, replicaPlacement *ReplicaPlacement, ttl *TTL) (v *Volume, e error) {
v = &Volume{dir: dirname, Collection: collection, Id: id}
v.SuperBlock = SuperBlock{ReplicaPlacement: replicaPlacement, Ttl: ttl}
v.needleMapKind = needleMapKind
e = v.load(true, true, needleMapKind)
return
}
func (v *Volume) String() string {
return fmt.Sprintf("Id:%v, dir:%s, Collection:%s, dataFile:%v, nm:%v, readOnly:%v", v.Id, v.dir, v.Collection, v.dataFile, v.nm, v.readOnly)
}
func loadVolumeWithoutIndex(dirname string, collection string, id VolumeId, needleMapKind NeedleMapType) (v *Volume, e error) {
v = &Volume{dir: dirname, Collection: collection, Id: id}
v.SuperBlock = SuperBlock{}
v.needleMapKind = needleMapKind
e = v.load(false, false, needleMapKind)
return
}
func (v *Volume) FileName() (fileName string) {
if v.Collection == "" {
fileName = path.Join(v.dir, v.Id.String())
} else {
fileName = path.Join(v.dir, v.Collection+"_"+v.Id.String())
}
return
}
func (v *Volume) load(alsoLoadIndex bool, createDatIfMissing bool, needleMapKind NeedleMapType) error {
var e error
fileName := v.FileName()
if exists, canRead, canWrite, modifiedTime := checkFile(fileName + ".dat"); exists {
if !canRead {
return fmt.Errorf("cannot read Volume Data file %s.dat", fileName)
}
if canWrite {
v.dataFile, e = os.OpenFile(fileName+".dat", os.O_RDWR|os.O_CREATE, 0644)
v.lastModifiedTime = uint64(modifiedTime.Unix())
} else {
glog.V(0).Infoln("opening " + fileName + ".dat in READONLY mode")
v.dataFile, e = os.Open(fileName + ".dat")
v.readOnly = true
}
} else {
if createDatIfMissing {
v.dataFile, e = os.OpenFile(fileName+".dat", os.O_RDWR|os.O_CREATE, 0644)
} else {
return fmt.Errorf("Volume Data file %s.dat does not exist.", fileName)
}
}
if e != nil {
if !os.IsPermission(e) {
return fmt.Errorf("cannot load Volume Data %s.dat: %v", fileName, e)
}
}
if v.ReplicaPlacement == nil {
e = v.readSuperBlock()
} else {
e = v.maybeWriteSuperBlock()
}
if e == nil && alsoLoadIndex {
var indexFile *os.File
if v.readOnly {
glog.V(1).Infoln("open to read file", fileName+".idx")
if indexFile, e = os.OpenFile(fileName+".idx", os.O_RDONLY, 0644); e != nil {
return fmt.Errorf("cannot read Volume Index %s.idx: %v", fileName, e)
}
} else {
glog.V(1).Infoln("open to write file", fileName+".idx")
if indexFile, e = os.OpenFile(fileName+".idx", os.O_RDWR|os.O_CREATE, 0644); e != nil {
return fmt.Errorf("cannot write Volume Index %s.idx: %v", fileName, e)
}
}
switch needleMapKind {
case NeedleMapInMemory:
glog.V(0).Infoln("loading index file", fileName+".idx", "readonly", v.readOnly)
if v.nm, e = LoadNeedleMap(indexFile); e != nil {
glog.V(0).Infof("loading index %s error: %v", fileName+".idx", e)
}
case NeedleMapLevelDb:
glog.V(0).Infoln("loading leveldb file", fileName+".ldb")
if v.nm, e = NewLevelDbNeedleMap(fileName+".ldb", indexFile); e != nil {
glog.V(0).Infof("loading leveldb %s error: %v", fileName+".ldb", e)
}
case NeedleMapBoltDb:
glog.V(0).Infoln("loading boltdb file", fileName+".bdb")
if v.nm, e = NewBoltDbNeedleMap(fileName+".bdb", indexFile); e != nil {
glog.V(0).Infof("loading boltdb %s error: %v", fileName+".bdb", e)
}
}
}
return e
}
func (v *Volume) Version() Version {
return v.SuperBlock.Version()
}
func (v *Volume) Size() int64 {
stat, e := v.dataFile.Stat()
if e == nil {
return stat.Size()
}
glog.V(0).Infof("Failed to read file size %s %v", v.dataFile.Name(), e)
return -1
}
// Close cleanly shuts down this volume
func (v *Volume) Close() {
v.accessLock.Lock()
defer v.accessLock.Unlock()
v.nm.Close()
_ = v.dataFile.Close()
}
func (v *Volume) NeedToReplicate() bool {
return v.ReplicaPlacement.GetCopyCount() > 1
}
// isFileUnchanged checks whether this needle to write is same as last one.
// It requires serialized access in the same volume.
func (v *Volume) isFileUnchanged(n *Needle) bool {
nv, ok := v.nm.Get(n.Id)
if ok && nv.Offset > 0 {
oldNeedle := new(Needle)
oldNeedle.Read(v.dataFile, int64(nv.Offset)*NeedlePaddingSize, nv.Size, v.Version())
if oldNeedle.Checksum == n.Checksum && bytes.Equal(oldNeedle.Data, n.Data) {
n.DataSize = oldNeedle.DataSize
return true
}
}
return false
}
// Destroy removes everything related to this volume
func (v *Volume) Destroy() (err error) {
if v.readOnly {
err = fmt.Errorf("%s is read-only", v.dataFile.Name())
return
}
v.Close()
err = os.Remove(v.dataFile.Name())
if err != nil {
return
}
err = v.nm.Destroy()
return
}
func (v *Volume) write(n *Needle) (size uint32, err error) {
glog.V(4).Infof("writing needle %s", NewFileIdFromNeedle(v.Id, n).String())
if v.readOnly {
err = fmt.Errorf("%s is read-only", v.dataFile.Name())
return
}
v.accessLock.Lock()
defer v.accessLock.Unlock()
if v.isFileUnchanged(n) {
size = n.DataSize
glog.V(4).Infof("needle is unchanged!")
return
}
var offset int64
if offset, err = v.dataFile.Seek(0, 2); err != nil {
glog.V(0).Infof("failed to seek the end of file: %v", err)
return
}
//ensure file writing starting from aligned positions
if offset%NeedlePaddingSize != 0 {
offset = offset + (NeedlePaddingSize - offset%NeedlePaddingSize)
if offset, err = v.dataFile.Seek(offset, 0); err != nil {
glog.V(0).Infof("failed to align in datafile %s: %v", v.dataFile.Name(), err)
return
}
}
if size, err = n.Append(v.dataFile, v.Version()); err != nil {
if e := v.dataFile.Truncate(offset); e != nil {
err = fmt.Errorf("%s\ncannot truncate %s: %v", err, v.dataFile.Name(), e)
}
return
}
nv, ok := v.nm.Get(n.Id)
if !ok || int64(nv.Offset)*NeedlePaddingSize < offset {
if err = v.nm.Put(n.Id, uint32(offset/NeedlePaddingSize), n.Size); err != nil {
glog.V(4).Infof("failed to save in needle map %d: %v", n.Id, err)
}
}
if v.lastModifiedTime < n.LastModified {
v.lastModifiedTime = n.LastModified
}
return
}
func (v *Volume) delete(n *Needle) (uint32, error) {
glog.V(4).Infof("delete needle %s", NewFileIdFromNeedle(v.Id, n).String())
if v.readOnly {
return 0, fmt.Errorf("%s is read-only", v.dataFile.Name())
}
v.accessLock.Lock()
defer v.accessLock.Unlock()
nv, ok := v.nm.Get(n.Id)
//fmt.Println("key", n.Id, "volume offset", nv.Offset, "data_size", n.Size, "cached size", nv.Size)
if ok {
size := nv.Size
if err := v.nm.Delete(n.Id); err != nil {
return size, err
}
if _, err := v.dataFile.Seek(0, 2); err != nil {
return size, err
}
n.Data = nil
_, err := n.Append(v.dataFile, v.Version())
return size, err
}
return 0, nil
}
func (v *Volume) read(n *Needle) (int, error) {
nv, ok := v.nm.Get(n.Id)
if !ok || nv.Offset == 0 {
return -1, errors.New("Not Found")
}
bytesRead, err := n.Read(v.dataFile, int64(nv.Offset)*NeedlePaddingSize, nv.Size, v.Version())
if err != nil {
return bytesRead, err
}
if !n.HasTtl() {
return bytesRead, err
}
ttlMinutes := n.Ttl.Minutes()
if ttlMinutes == 0 {
return bytesRead, nil
}
if !n.HasLastModifiedDate() {
return bytesRead, nil
}
if uint64(time.Now().Unix()) < n.LastModified+uint64(ttlMinutes*60) {
return bytesRead, nil
}
return -1, errors.New("Not Found")
}
func ScanVolumeFile(dirname string, collection string, id VolumeId,
needleMapKind NeedleMapType,
visitSuperBlock func(SuperBlock) error,
readNeedleBody bool,
visitNeedle func(n *Needle, offset int64) error) (err error) {
var v *Volume
if v, err = loadVolumeWithoutIndex(dirname, collection, id, needleMapKind); err != nil {
return fmt.Errorf("Failed to load volume %d: %v", id, err)
}
if err = visitSuperBlock(v.SuperBlock); err != nil {
return fmt.Errorf("Failed to read volume %d super block: %v", id, err)
}
version := v.Version()
offset := int64(SuperBlockSize)
n, rest, e := ReadNeedleHeader(v.dataFile, version, offset)
if e != nil {
err = fmt.Errorf("cannot read needle header: %v", e)
return
}
for n != nil {
if readNeedleBody {
if err = n.ReadNeedleBody(v.dataFile, version, offset+int64(NeedleHeaderSize), rest); err != nil {
glog.V(0).Infof("cannot read needle body: %v", err)
//err = fmt.Errorf("cannot read needle body: %v", err)
//return
}
if n.DataSize >= n.Size {
// this should come from a bug reported on #87 and #93
// fixed in v0.69
// remove this whole "if" clause later, long after 0.69
oldRest, oldSize := rest, n.Size
padding := NeedlePaddingSize - ((n.Size + NeedleHeaderSize + NeedleChecksumSize) % NeedlePaddingSize)
n.Size = 0
rest = n.Size + NeedleChecksumSize + padding
if rest%NeedlePaddingSize != 0 {
rest += (NeedlePaddingSize - rest%NeedlePaddingSize)
}
glog.V(4).Infof("Adjusting n.Size %d=>0 rest:%d=>%d %+v", oldSize, oldRest, rest, n)
}
}
if err = visitNeedle(n, offset); err != nil {
glog.V(0).Infof("visit needle error: %v", err)
}
offset += int64(NeedleHeaderSize) + int64(rest)
glog.V(4).Infof("==> new entry offset %d", offset)
if n, rest, err = ReadNeedleHeader(v.dataFile, version, offset); err != nil {
if err == io.EOF {
return nil
}
return fmt.Errorf("cannot read needle header: %v", err)
}
glog.V(4).Infof("new entry needle size:%d rest:%d", n.Size, rest)
}
return
}
func (v *Volume) ContentSize() uint64 {
return v.nm.ContentSize()
}
func checkFile(filename string) (exists, canRead, canWrite bool, modTime time.Time) {
exists = true
fi, err := os.Stat(filename)
if os.IsNotExist(err) {
exists = false
return
}
if fi.Mode()&0400 != 0 {
canRead = true
}
if fi.Mode()&0200 != 0 {
canWrite = true
}
modTime = fi.ModTime()
return
}
// volume is expired if modified time + volume ttl < now
// except when volume is empty
// or when the volume does not have a ttl
// or when volumeSizeLimit is 0 when server just starts
func (v *Volume) expired(volumeSizeLimit uint64) bool {
if volumeSizeLimit == 0 {
//skip if we don't know size limit
return false
}
if v.ContentSize() == 0 {
return false
}
if v.Ttl == nil || v.Ttl.Minutes() == 0 {
return false
}
glog.V(0).Infof("now:%v lastModified:%v", time.Now().Unix(), v.lastModifiedTime)
livedMinutes := (time.Now().Unix() - int64(v.lastModifiedTime)) / 60
glog.V(0).Infof("ttl:%v lived:%v", v.Ttl, livedMinutes)
if int64(v.Ttl.Minutes()) < livedMinutes {
return true
}
return false
}
// wait either maxDelayMinutes or 10% of ttl minutes
func (v *Volume) exiredLongEnough(maxDelayMinutes uint32) bool {
if v.Ttl == nil || v.Ttl.Minutes() == 0 {
return false
}
removalDelay := v.Ttl.Minutes() / 10
if removalDelay > maxDelayMinutes {
removalDelay = maxDelayMinutes
}
if uint64(v.Ttl.Minutes()+removalDelay)*60+v.lastModifiedTime < uint64(time.Now().Unix()) {
return true
}
return false
}