seaweedfs/weed/util/log_buffer/log_buffer.go
2024-01-05 15:35:42 -08:00

327 lines
8.3 KiB
Go

package log_buffer
import (
"bytes"
"sync"
"sync/atomic"
"time"
"google.golang.org/protobuf/proto"
"github.com/seaweedfs/seaweedfs/weed/glog"
"github.com/seaweedfs/seaweedfs/weed/pb/filer_pb"
"github.com/seaweedfs/seaweedfs/weed/util"
)
const BufferSize = 4 * 1024 * 1024
const PreviousBufferCount = 3
type dataToFlush struct {
startTime time.Time
stopTime time.Time
data *bytes.Buffer
}
type LogBuffer struct {
name string
prevBuffers *SealedBuffers
buf []byte
idx []int
pos int
startTime time.Time
stopTime time.Time
lastFlushTime time.Time
sizeBuf []byte
flushInterval time.Duration
flushFn func(startTime, stopTime time.Time, buf []byte)
notifyFn func()
isStopping *atomic.Bool
flushChan chan *dataToFlush
lastTsNs int64
sync.RWMutex
}
func NewLogBuffer(name string, flushInterval time.Duration, flushFn func(startTime, stopTime time.Time, buf []byte), notifyFn func()) *LogBuffer {
lb := &LogBuffer{
name: name,
prevBuffers: newSealedBuffers(PreviousBufferCount),
buf: make([]byte, BufferSize),
sizeBuf: make([]byte, 4),
flushInterval: flushInterval,
flushFn: flushFn,
notifyFn: notifyFn,
flushChan: make(chan *dataToFlush, 256),
isStopping: new(atomic.Bool),
}
go lb.loopFlush()
go lb.loopInterval()
return lb
}
func (m *LogBuffer) AddToBuffer(partitionKey, data []byte, processingTsNs int64) {
var toFlush *dataToFlush
m.Lock()
defer func() {
m.Unlock()
if toFlush != nil {
m.flushChan <- toFlush
}
if m.notifyFn != nil {
m.notifyFn()
}
}()
// need to put the timestamp inside the lock
var ts time.Time
if processingTsNs == 0 {
ts = time.Now()
processingTsNs = ts.UnixNano()
} else {
ts = time.Unix(0, processingTsNs)
}
if m.lastTsNs >= processingTsNs {
// this is unlikely to happen, but just in case
processingTsNs = m.lastTsNs + 1
ts = time.Unix(0, processingTsNs)
}
m.lastTsNs = processingTsNs
logEntry := &filer_pb.LogEntry{
TsNs: processingTsNs,
PartitionKeyHash: util.HashToInt32(partitionKey),
Data: data,
}
logEntryData, _ := proto.Marshal(logEntry)
size := len(logEntryData)
if m.pos == 0 {
m.startTime = ts
}
if m.startTime.Add(m.flushInterval).Before(ts) || len(m.buf)-m.pos < size+4 {
// glog.V(4).Infof("%s copyToFlush1 start time %v, ts %v, remaining %d bytes", m.name, m.startTime, ts, len(m.buf)-m.pos)
toFlush = m.copyToFlush()
m.startTime = ts
if len(m.buf) < size+4 {
m.buf = make([]byte, 2*size+4)
}
}
m.stopTime = ts
m.idx = append(m.idx, m.pos)
util.Uint32toBytes(m.sizeBuf, uint32(size))
copy(m.buf[m.pos:m.pos+4], m.sizeBuf)
copy(m.buf[m.pos+4:m.pos+4+size], logEntryData)
m.pos += size + 4
// fmt.Printf("partitionKey %v entry size %d total %d count %d\n", string(partitionKey), size, m.pos, len(m.idx))
}
func (m *LogBuffer) IsStopping() bool {
return m.isStopping.Load()
}
func (m *LogBuffer) Shutdown() {
isAlreadyStopped := m.isStopping.Swap(true)
if isAlreadyStopped {
return
}
toFlush := m.copyToFlush()
m.flushChan <- toFlush
close(m.flushChan)
}
func (m *LogBuffer) loopFlush() {
for d := range m.flushChan {
if d != nil {
// glog.V(4).Infof("%s flush [%v, %v] size %d", m.name, d.startTime, d.stopTime, len(d.data.Bytes()))
m.flushFn(d.startTime, d.stopTime, d.data.Bytes())
d.releaseMemory()
// local logbuffer is different from aggregate logbuffer here
m.lastFlushTime = d.stopTime
}
}
}
func (m *LogBuffer) loopInterval() {
for !m.IsStopping() {
time.Sleep(m.flushInterval)
if m.IsStopping() {
return
}
m.Lock()
toFlush := m.copyToFlush()
m.Unlock()
if toFlush != nil {
glog.V(0).Infof("%s flush [%v, %v] size %d", m.name, toFlush.startTime, toFlush.stopTime, len(toFlush.data.Bytes()))
m.flushChan <- toFlush
} else {
// glog.V(0).Infof("%s no flush", m.name)
}
}
}
func (m *LogBuffer) copyToFlush() *dataToFlush {
if m.pos > 0 {
// fmt.Printf("flush buffer %d pos %d empty space %d\n", len(m.buf), m.pos, len(m.buf)-m.pos)
var d *dataToFlush
if m.flushFn != nil {
d = &dataToFlush{
startTime: m.startTime,
stopTime: m.stopTime,
data: copiedBytes(m.buf[:m.pos]),
}
// glog.V(4).Infof("%s flushing [0,%d) with %d entries [%v, %v]", m.name, m.pos, len(m.idx), m.startTime, m.stopTime)
} else {
// glog.V(4).Infof("%s removed from memory [0,%d) with %d entries [%v, %v]", m.name, m.pos, len(m.idx), m.startTime, m.stopTime)
m.lastFlushTime = m.stopTime
}
m.buf = m.prevBuffers.SealBuffer(m.startTime, m.stopTime, m.buf, m.pos)
m.startTime = time.Unix(0, 0)
m.stopTime = time.Unix(0, 0)
m.pos = 0
m.idx = m.idx[:0]
return d
}
return nil
}
func (d *dataToFlush) releaseMemory() {
d.data.Reset()
bufferPool.Put(d.data)
}
func (m *LogBuffer) ReadFromBuffer(lastReadTime time.Time) (bufferCopy *bytes.Buffer, err error) {
m.RLock()
defer m.RUnlock()
// Read from disk and memory
// 1. read from disk, last time is = td
// 2. in memory, the earliest time = tm
// if tm <= td, case 2.1
// read from memory
// if tm is empty, case 2.2
// read from memory
// if td < tm, case 2.3
// read from disk again
var tsMemory time.Time
if !m.startTime.IsZero() {
tsMemory = m.startTime
}
for _, prevBuf := range m.prevBuffers.buffers {
if !prevBuf.startTime.IsZero() && prevBuf.startTime.Before(tsMemory) {
tsMemory = prevBuf.startTime
}
}
if tsMemory.IsZero() { // case 2.2
return nil, nil
} else if lastReadTime.Before(tsMemory) { // case 2.3
if !m.lastFlushTime.IsZero() {
glog.V(0).Infof("resume with last flush time: %v", m.lastFlushTime)
return nil, ResumeFromDiskError
}
}
// the following is case 2.1
if lastReadTime.Equal(m.stopTime) {
return nil, nil
}
if lastReadTime.After(m.stopTime) {
// glog.Fatalf("unexpected last read time %v, older than latest %v", lastReadTime, m.stopTime)
return nil, nil
}
if lastReadTime.Before(m.startTime) {
// println("checking ", lastReadTime.UnixNano())
for _, buf := range m.prevBuffers.buffers {
if buf.startTime.After(lastReadTime) {
// glog.V(4).Infof("%s return the %d sealed buffer %v", m.name, i, buf.startTime)
// println("return the", i, "th in memory", buf.startTime.UnixNano())
return copiedBytes(buf.buf[:buf.size]), nil
}
if !buf.startTime.After(lastReadTime) && buf.stopTime.After(lastReadTime) {
pos := buf.locateByTs(lastReadTime)
// fmt.Printf("locate buffer[%d] pos %d\n", i, pos)
return copiedBytes(buf.buf[pos:buf.size]), nil
}
}
// glog.V(4).Infof("%s return the current buf %v", m.name, lastReadTime)
return copiedBytes(m.buf[:m.pos]), nil
}
lastTs := lastReadTime.UnixNano()
l, h := 0, len(m.idx)-1
/*
for i, pos := range m.idx {
logEntry, ts := readTs(m.buf, pos)
event := &filer_pb.SubscribeMetadataResponse{}
proto.Unmarshal(logEntry.Data, event)
entry := event.EventNotification.OldEntry
if entry == nil {
entry = event.EventNotification.NewEntry
}
fmt.Printf("entry %d ts: %v offset:%d dir:%s name:%s\n", i, time.Unix(0, ts), pos, event.Directory, entry.Name)
}
fmt.Printf("l=%d, h=%d\n", l, h)
*/
for l <= h {
mid := (l + h) / 2
pos := m.idx[mid]
_, t := readTs(m.buf, pos)
if t <= lastTs {
l = mid + 1
} else if lastTs < t {
var prevT int64
if mid > 0 {
_, prevT = readTs(m.buf, m.idx[mid-1])
}
if prevT <= lastTs {
// fmt.Printf("found l=%d, m-1=%d(ts=%d), m=%d(ts=%d), h=%d [%d, %d) \n", l, mid-1, prevT, mid, t, h, pos, m.pos)
return copiedBytes(m.buf[pos:m.pos]), nil
}
h = mid
}
// fmt.Printf("l=%d, h=%d\n", l, h)
}
// FIXME: this could be that the buffer has been flushed already
return nil, nil
}
func (m *LogBuffer) ReleaseMemory(b *bytes.Buffer) {
bufferPool.Put(b)
}
var bufferPool = sync.Pool{
New: func() interface{} {
return new(bytes.Buffer)
},
}
func copiedBytes(buf []byte) (copied *bytes.Buffer) {
copied = bufferPool.Get().(*bytes.Buffer)
copied.Reset()
copied.Write(buf)
return
}
func readTs(buf []byte, pos int) (size int, ts int64) {
size = int(util.BytesToUint32(buf[pos : pos+4]))
entryData := buf[pos+4 : pos+4+size]
logEntry := &filer_pb.LogEntry{}
err := proto.Unmarshal(entryData, logEntry)
if err != nil {
glog.Fatalf("unexpected unmarshal filer_pb.LogEntry: %v", err)
}
return size, logEntry.TsNs
}