使用Go语言，25秒读取16GB文件
点击上方“程序员大白”，选择“星标”公众号
重磅干货，第一时间送达
原文链接：https://medium.com/swlh/processing-16gb-file-in-seconds-go-lang-3982c235dfa2
文章转载：分布式实验室
当今世界的任何计算机系统每天都会生成大量的日志或数据。随着系统的发展，将调试数据存储到数据库中是不可行的，因为它们是不可变的，并且只能用于分析和解决故障。所以大部分公司倾向于将日志存储在文件中，而这些文件通常位于本地磁盘中。
让我们开始编码……
首先，我们打开文件。对于任何文件的IO，我们都将使用标准的Go os.File。
f, err := os.Open(fileName)
if err != nil {
fmt.Println("cannot able to read the file", err)
return
}
// UPDATE: close after checking error
defer file.Close() //Do not forget to close the file
打开文件后，我们有以下两个选项可以选择：
逐行读取文件，这有助于减少内存紧张，但需要更多的时间。一次将整个文件读入内存并处理该文件，这将消耗更多内存，但会显著减少时间。
由于文件太大，即16 GB，因此无法将整个文件加载到内存中。但是第一种选择对我们来说也是不可行的，因为我们希望在几秒钟内处理文件。
但你猜怎么着，还有第三种选择。瞧……相比于将整个文件加载到内存中，在Go语言中，我们还可以使用bufio.NewReader()将文件分块加载。
r := bufio.NewReader(f)
for {
buf := make([]byte,4*1024) //the chunk size
n, err := r.Read(buf) //loading chunk into buffer
buf = buf[:n]
if n == 0 {
if err != nil {
fmt.Println(err)
break
}
if err == io.EOF {
break
}
return err
}
}
一旦我们将文件分块，我们就可以分叉一个线程，即Go routine，同时处理多个文件区块。上述代码将修改为：
//sync pools to reuse the memory and decrease the preassure on Garbage Collector
linesPool := sync.Pool{New: func() interface{} {
lines := make([]byte, 500*1024)
return lines
}}
stringPool := sync.Pool{New: func() interface{} {
lines := ""
return lines
}}
slicePool := sync.Pool{New: func() interface{} {
lines := make([]string, 100)
return lines
}}
r := bufio.NewReader(f)
var wg sync.WaitGroup //wait group to keep track off all threads
for {
buf := linesPool.Get().([]byte)
n, err := r.Read(buf)
buf = buf[:n]
if n == 0 {
if err != nil {
fmt.Println(err)
break
}
if err == io.EOF {
break
}
return err
}
nextUntillNewline, err := r.ReadBytes('\n')//read entire line
if err != io.EOF {
buf = append(buf, nextUntillNewline...)
}
wg.Add(1)
go func() {
//process each chunk concurrently
//start -> log start time, end -> log end time
ProcessChunk(buf, &linesPool, &stringPool, &slicePool, start, end)
wg.Done()
}()
}
wg.Wait()
}
上面的代码，引入了两个优化点：
sync.Pool是一个强大的对象池，可以重用对象来减轻垃圾收集器的压力。我们将重用各个分片的内存，以减少内存消耗，大大加快我们的工作。Go Routines帮助我们同时处理缓冲区块，这大大提高了处理速度。
现在让我们实现ProcessChunk函数，它将处理以下格式的日志行。
2020-01-31T20:12:38.1234Z, Some Field, Other Field, And so on, Till new line,...\n
我们将根据命令行提供的时间戳提取日志。
func ProcessChunk(chunk []byte, linesPool *sync.Pool, stringPool *sync.Pool, slicePool *sync.Pool, start time.Time, end time.Time) {
//another wait group to process every chunk further
var wg2 sync.WaitGroup
logs := stringPool.Get().(string)
logs = string(chunk)
linesPool.Put(chunk) //put back the chunk in pool
//split the string by "\n", so that we have slice of logs
logsSlice := strings.Split(logs, "\n")
stringPool.Put(logs) //put back the string pool
chunkSize := 100 //process the bunch of 100 logs in thread
n := len(logsSlice)
noOfThread := n / chunkSize
if n%chunkSize != 0 { //check for overflow
noOfThread++
}
length := len(logsSlice)
//traverse the chunk
for i := 0; i < length; i += chunkSize {
wg2.Add(1)
//process each chunk in saperate chunk
go func(s int, e int) {
for i:= s; i<e;i++{
text := logsSlice[i]
if len(text) == 0 {
continue
}
logParts := strings.SplitN(text, ",", 2)
logCreationTimeString := logParts[0]
logCreationTime, err := time.Parse("2006-01- 02T15:04:05.0000Z", logCreationTimeString)
if err != nil {
fmt.Printf("\n Could not able to parse the time :%s for log : %v", logCreationTimeString, text)
return
}
// check if log's timestamp is inbetween our desired period
if logCreationTime.After(start) && logCreationTime.Before(end) {
fmt.Println(text)
}
}
textSlice = nil
wg2.Done()
}(i*chunkSize, int(math.Min(float64((i+1)*chunkSize), float64(len(logsSlice)))))
//passing the indexes for processing
}
wg2.Wait() //wait for a chunk to finish
logsSlice = nil
}
对上面的代码进行基准测试。以16 GB的日志文件为例，提取日志所需的时间约为25秒。
完整的代码示例如下：
func main() {
s := time.Now()
args := os.Args[1:]
if len(args) != 6 { // for format LogExtractor.exe -f "From Time" -t "To Time" -i "Log file directory location"
fmt.Println("Please give proper command line arguments")
return
}
startTimeArg := args[1]
finishTimeArg := args[3]
fileName := args[5]
file, err := os.Open(fileName)
if err != nil {
fmt.Println("cannot able to read the file", err)
return
}
defer file.Close() //close after checking err
queryStartTime, err := time.Parse("2006-01-02T15:04:05.0000Z", startTimeArg)
if err != nil {
fmt.Println("Could not able to parse the start time", startTimeArg)
return
}
queryFinishTime, err := time.Parse("2006-01-02T15:04:05.0000Z", finishTimeArg)
if err != nil {
fmt.Println("Could not able to parse the finish time", finishTimeArg)
return
}
filestat, err := file.Stat()
if err != nil {
fmt.Println("Could not able to get the file stat")
return
}
fileSize := filestat.Size()
offset := fileSize - 1
lastLineSize := 0
for {
b := make([]byte, 1)
n, err := file.ReadAt(b, offset)
if err != nil {
fmt.Println("Error reading file ", err)
break
}
char := string(b[0])
if char == "\n" {
break
}
offset--
lastLineSize += n
}
lastLine := make([]byte, lastLineSize)
_, err = file.ReadAt(lastLine, offset+1)
if err != nil {
fmt.Println("Could not able to read last line with offset", offset, "and lastline size", lastLineSize)
return
}
logSlice := strings.SplitN(string(lastLine), ",", 2)
logCreationTimeString := logSlice[0]
lastLogCreationTime, err := time.Parse("2006-01-02T15:04:05.0000Z", logCreationTimeString)
if err != nil {
fmt.Println("can not able to parse time : ", err)
}
if lastLogCreationTime.After(queryStartTime) && lastLogCreationTime.Before(queryFinishTime) {
Process(file, queryStartTime, queryFinishTime)
}
fmt.Println("\nTime taken - ", time.Since(s))
}
func Process(f *os.File, start time.Time, end time.Time) error {
linesPool := sync.Pool{New: func() interface{} {
lines := make([]byte, 250*1024)
return lines
}}
stringPool := sync.Pool{New: func() interface{} {
lines := ""
return lines
}}
r := bufio.NewReader(f)
var wg sync.WaitGroup
for {
buf := linesPool.Get().([]byte)
n, err := r.Read(buf)
buf = buf[:n]
if n == 0 {
if err != nil {
fmt.Println(err)
break
}
if err == io.EOF {
break
}
return err
}
nextUntillNewline, err := r.ReadBytes('\n')
if err != io.EOF {
buf = append(buf, nextUntillNewline...)
}
wg.Add(1)
go func() {
ProcessChunk(buf, &linesPool, &stringPool, start, end)
wg.Done()
}()
}
wg.Wait()
return nil
}
func ProcessChunk(chunk []byte, linesPool *sync.Pool, stringPool *sync.Pool, start time.Time, end time.Time) {
var wg2 sync.WaitGroup
logs := stringPool.Get().(string)
logs = string(chunk)
linesPool.Put(chunk)
logsSlice := strings.Split(logs, "\n")
stringPool.Put(logs)
chunkSize := 300
n := len(logsSlice)
noOfThread := n / chunkSize
if n%chunkSize != 0 {
noOfThread++
}
for i := 0; i < (noOfThread); i++ {
wg2.Add(1)
go func(s int, e int) {
defer wg2.Done() //to avaoid deadlocks
for i := s; i < e; i++ {
text := logsSlice[i]
if len(text) == 0 {
continue
}
logSlice := strings.SplitN(text, ",", 2)
logCreationTimeString := logSlice[0]
logCreationTime, err := time.Parse("2006-01-02T15:04:05.0000Z", logCreationTimeString)
if err != nil {
fmt.Printf("\n Could not able to parse the time :%s for log : %v", logCreationTimeString, text)
return
}
if logCreationTime.After(start) && logCreationTime.Before(end) {
//fmt.Println(text)
}
}
}(i*chunkSize, int(math.Min(float64((i+1)*chunkSize), float64(len(logsSlice)))))
}
wg2.Wait()
logsSlice = nil
}推荐阅读
关于程序员大白
程序员大白是一群哈工大，东北大学，西湖大学和上海交通大学的硕士博士运营维护的号，大家乐于分享高质量文章，喜欢总结知识，欢迎关注[程序员大白]，大家一起学习进步！
评论