gitlab-com-seaweedfs/seaweedfs
1
0
Fork 0
mirror of https://github.com/seaweedfs/seaweedfs.git synced 2024-01-19 02:48:24 +00:00
Code Issues Projects Releases Wiki Activity

Merge pull request #3337 from guol-fnst/loading_volume

Browse source
This commit is contained in:
Chris Lu 2022-07-20 09:06:06 -07:00 committed by GitHub
parent efec31de8d ac694f0c8f
commit 9ec0d1caaa
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
7 changed files with 93 additions and 20 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
unmaintained/diff_volume_servers/diff_volume_servers.go
View file

@ -6,6 +6,10 @@ import (
"errors"
"flag"
"fmt"
"io"
"math"
"os"
"github.com/chrislusf/seaweedfs/weed/glog"
"github.com/chrislusf/seaweedfs/weed/operation"
"github.com/chrislusf/seaweedfs/weed/pb"
@ -16,9 +20,6 @@ import (
"github.com/chrislusf/seaweedfs/weed/storage/types"
"github.com/chrislusf/seaweedfs/weed/util"
"google.golang.org/grpc"
"io"
"math"
"os"
)
var (
@ -155,7 +156,7 @@ func getVolumeFiles(v uint32, addr pb.ServerAddress) (map[types.NeedleId]needleS
var maxOffset int64
files := map[types.NeedleId]needleState{}
err = idx.WalkIndexFile(idxFile, func(key types.NeedleId, offset types.Offset, size types.Size) error {
err = idx.WalkIndexFile(idxFile, 0, func(key types.NeedleId, offset types.Offset, size types.Size) error {
if offset.IsZero() || size.IsDeleted() {
files[key] = needleState{
state: stateDeleted,

5
unmaintained/see_idx/see_idx.go
View file

@ -3,11 +3,12 @@ package main
import (
"flag"
"fmt"
"github.com/chrislusf/seaweedfs/weed/util"
"os"
"path"
"strconv"
"github.com/chrislusf/seaweedfs/weed/util"
"github.com/chrislusf/seaweedfs/weed/glog"
"github.com/chrislusf/seaweedfs/weed/storage/idx"
"github.com/chrislusf/seaweedfs/weed/storage/types"
@ -36,7 +37,7 @@ func main() {
}
defer indexFile.Close()
idx.WalkIndexFile(indexFile, func(key types.NeedleId, offset types.Offset, size types.Size) error {
idx.WalkIndexFile(indexFile, 0, func(key types.NeedleId, offset types.Offset, size types.Size) error {
fmt.Printf("key:%v offset:%v size:%v(%v)\n", key, offset, size, util.BytesToHumanReadable(uint64(size)))
return nil
})

2
weed/storage/erasure_coding/ec_encoder.go
View file

@ -294,7 +294,7 @@ func readNeedleMap(baseFileName string) (*needle_map.MemDb, error) {
defer indexFile.Close()
cm := needle_map.NewMemDb()
err = idx.WalkIndexFile(indexFile, func(key types.NeedleId, offset types.Offset, size types.Size) error {
err = idx.WalkIndexFile(indexFile, 0, func(key types.NeedleId, offset types.Offset, size types.Size) error {
if !offset.IsZero() && size != types.TombstoneFileSize {
cm.Set(key, offset, size)
} else {

4
weed/storage/idx/walk.go
View file

@ -9,8 +9,8 @@ import (
// walks through the index file, calls fn function with each key, offset, size
// stops with the error returned by the fn function
func WalkIndexFile(r io.ReaderAt, fn func(key types.NeedleId, offset types.Offset, size types.Size) error) error {
var readerOffset int64
func WalkIndexFile(r io.ReaderAt, startFrom uint64, fn func(key types.NeedleId, offset types.Offset, size types.Size) error) error {
readerOffset := int64(startFrom * types.NeedleMapEntrySize)
bytes := make([]byte, types.NeedleMapEntrySize*RowsToRead)
count, e := r.ReadAt(bytes, readerOffset)
if count == 0 && e == io.EOF {

2
weed/storage/needle_map/memdb.go
View file

@ -111,7 +111,7 @@ func (cm *MemDb) LoadFromIdx(idxName string) (ret error) {
func (cm *MemDb) LoadFromReaderAt(readerAt io.ReaderAt) (ret error) {
return idx.WalkIndexFile(readerAt, func(key NeedleId, offset Offset, size Size) error {
return idx.WalkIndexFile(readerAt, 0, func(key NeedleId, offset Offset, size Size) error {
if offset.IsZero() || size.IsDeleted() {
return cm.Delete(key)
}

85
weed/storage/needle_map_leveldb.go
View file

@ -9,6 +9,8 @@ import (
"github.com/syndtr/goleveldb/leveldb/opt"
"github.com/chrislusf/seaweedfs/weed/storage/idx"
"github.com/chrislusf/seaweedfs/weed/storage/types"
"github.com/chrislusf/seaweedfs/weed/util"
"github.com/syndtr/goleveldb/leveldb"
@ -17,10 +19,16 @@ import (
. "github.com/chrislusf/seaweedfs/weed/storage/types"
)
//mark it every watermarkBatchSize operations
const watermarkBatchSize = 10000
var watermarkKey = []byte("idx_entry_watermark")
type LevelDbNeedleMap struct {
baseNeedleMapper
dbFileName string
db *leveldb.DB
recordCount uint64
}
func NewLevelDbNeedleMap(dbFileName string, indexFile *os.File, opts *opt.Options) (m *LevelDbNeedleMap, err error) {
@ -46,7 +54,14 @@ func NewLevelDbNeedleMap(dbFileName string, indexFile *os.File, opts *opt.Option
return
}
}
glog.V(1).Infof("Loading %s...", indexFile.Name())
glog.V(0).Infof("Loading %s... , watermark: %d", dbFileName, getWatermark(m.db))
m.recordCount = uint64(m.indexFileOffset / types.NeedleMapEntrySize)
watermark := (m.recordCount / watermarkBatchSize) * watermarkBatchSize
err = setWatermark(m.db, watermark)
if err != nil {
glog.Fatalf("set watermark for %s error: %s\n", dbFileName, err)
return
}
mm, indexLoadError := newNeedleMapMetricFromIndexFile(indexFile)
if indexLoadError != nil {
return nil, indexLoadError
@ -78,9 +93,20 @@ func generateLevelDbFile(dbFileName string, indexFile *os.File) error {
return err
}
defer db.Close()
return idx.WalkIndexFile(indexFile, func(key NeedleId, offset Offset, size Size) error {
watermark := getWatermark(db)
if stat, err := indexFile.Stat(); err != nil {
glog.Fatalf("stat file %s: %v", indexFile.Name(), err)
return err
} else {
if watermark*types.NeedleMapEntrySize > uint64(stat.Size()) {
glog.Warningf("wrong watermark %d for filesize %d", watermark, stat.Size())
}
glog.V(0).Infof("generateLevelDbFile %s, watermark %d, num of entries:%d", dbFileName, watermark, (uint64(stat.Size())-watermark*types.NeedleMapEntrySize)/types.NeedleMapEntrySize)
}
return idx.WalkIndexFile(indexFile, watermark, func(key NeedleId, offset Offset, size Size) error {
if !offset.IsZero() && size.IsValid() {
levelDbWrite(db, key, offset, size)
levelDbWrite(db, key, offset, size, false, 0)
} else {
levelDbDelete(db, key)
}
@ -102,6 +128,7 @@ func (m *LevelDbNeedleMap) Get(key NeedleId) (element *needle_map.NeedleValue, o
func (m *LevelDbNeedleMap) Put(key NeedleId, offset Offset, size Size) error {
var oldSize Size
var watermark uint64
if oldNeedle, ok := m.Get(key); ok {
oldSize = oldNeedle.Size
}
@ -110,16 +137,54 @@ func (m *LevelDbNeedleMap) Put(key NeedleId, offset Offset, size Size) error {
if err := m.appendToIndexFile(key, offset, size); err != nil {
return fmt.Errorf("cannot write to indexfile %s: %v", m.indexFile.Name(), err)
}
return levelDbWrite(m.db, key, offset, size)
m.recordCount++
if m.recordCount%watermarkBatchSize != 0 {
watermark = 0
} else {
watermark = (m.recordCount / watermarkBatchSize) * watermarkBatchSize
glog.V(1).Infof("put cnt:%d for %s,watermark: %d", m.recordCount, m.dbFileName, watermark)
}
return levelDbWrite(m.db, key, offset, size, watermark == 0, watermark)
}
func levelDbWrite(db *leveldb.DB, key NeedleId, offset Offset, size Size) error {
func getWatermark(db *leveldb.DB) uint64 {
data, err := db.Get(watermarkKey, nil)
if err != nil || len(data) != 8 {
glog.Warningf("get watermark from db error: %v, %d", err, len(data))
/*
if !strings.Contains(strings.ToLower(err.Error()), "not found") {
err = setWatermark(db, 0)
if err != nil {
glog.Errorf("failed to set watermark: %v", err)
}
}
*/
return 0
}
return util.BytesToUint64(data)
}
func setWatermark(db *leveldb.DB, watermark uint64) error {
glog.V(1).Infof("set watermark %d", watermark)
var wmBytes = make([]byte, 8)
util.Uint64toBytes(wmBytes, watermark)
if err := db.Put(watermarkKey, wmBytes, nil); err != nil {
return fmt.Errorf("failed to setWatermark: %v", err)
}
return nil
}
func levelDbWrite(db *leveldb.DB, key NeedleId, offset Offset, size Size, updateWatermark bool, watermark uint64) error {
bytes := needle_map.ToBytes(key, offset, size)
if err := db.Put(bytes[0:NeedleIdSize], bytes[NeedleIdSize:NeedleIdSize+OffsetSize+SizeSize], nil); err != nil {
return fmt.Errorf("failed to write leveldb: %v", err)
}
// set watermark
if updateWatermark {
return setWatermark(db, watermark)
}
return nil
}
func levelDbDelete(db *leveldb.DB, key NeedleId) error {
@ -129,6 +194,7 @@ func levelDbDelete(db *leveldb.DB, key NeedleId) error {
}
func (m *LevelDbNeedleMap) Delete(key NeedleId, offset Offset) error {
var watermark uint64
oldNeedle, found := m.Get(key)
if !found || oldNeedle.Size.IsDeleted() {
return nil
@ -139,8 +205,13 @@ func (m *LevelDbNeedleMap) Delete(key NeedleId, offset Offset) error {
if err := m.appendToIndexFile(key, offset, TombstoneFileSize); err != nil {
return err
}
return levelDbWrite(m.db, key, oldNeedle.Offset, -oldNeedle.Size)
m.recordCount++
if m.recordCount%watermarkBatchSize != 0 {
watermark = 0
} else {
watermark = (m.recordCount / watermarkBatchSize) * watermarkBatchSize
}
return levelDbWrite(m.db, key, oldNeedle.Offset, -oldNeedle.Size, watermark == 0, watermark)
}
func (m *LevelDbNeedleMap) Close() {

2
weed/storage/needle_map_memory.go
View file

@ -33,7 +33,7 @@ func LoadCompactNeedleMap(file *os.File) (*NeedleMap, error) {
}
func doLoading(file *os.File, nm *NeedleMap) (*NeedleMap, error) {
e := idx.WalkIndexFile(file, func(key NeedleId, offset Offset, size Size) error {
e := idx.WalkIndexFile(file, 0, func(key NeedleId, offset Offset, size Size) error {
nm.MaybeSetMaxFileKey(key)
if !offset.IsZero() && size.IsValid() {
nm.FileCounter++