Go實現海量日誌收集系統(三)

時間  2021-02-14
標籤 html python linux nginx git github 算法 服務器 restful app 欄目 系統性能 简体版
原文   原文鏈接

51Reboot 將在 2020.1.16日 21：00 爲您帶來分享主題《大佬教你如何從 ES 初學者到 ES專家》html

直播連接（提早報名）：https://ke.qq.com/course/482014?taid=4309905192737502&tuin=31589b0epython

再次整理了一下這個日誌收集系統的框，以下圖linux

Go實現海量日誌收集系統(三)

此次要實現的代碼的總體邏輯爲：nginx

Go實現海量日誌收集系統(三)

完整代碼地址爲: https://github.com/pythonsite/logagentgit

etcd介紹

高可用的分佈式key-value存儲，能夠用於配置共享和服務發現github

相似的項目：zookeeper和consul算法

開發語言：go服務器

接口：提供restful的接口，使用簡單restful

實現算法：基於raft算法的強一致性，高可用的服務存儲目錄app

etcd的應用場景：

  • 服務發現和服務註冊
  • 配置中心(咱們實現的日誌收集客戶端須要用到)
  • 分佈式鎖
  • master選舉

官網對etcd的有一個很是簡明的介紹：

Go實現海量日誌收集系統(三)

etcd搭建：
下載地址：https://github.com/coreos/etcd/releases/
根據本身的環境下載對應的版本而後啓動起來就能夠了

啓動以後能夠經過以下命令驗證一下：

[root@localhost etcd-v3.2.18-linux-amd64]# ./etcdctl set name zhaofan 

zhaofan
[root@localhost etcd-v3.2.18-linux-amd64]# ./etcdctl get name
zhaofan
[root@localhost etcd-v3.2.18-linux-amd64]#

context 介紹和使用

其實這個東西翻譯過來就是上下文管理，那麼context的做用是作什麼，主要有以下兩個做用：

  • 控制goroutine的超時
  • 保存上下文數據

經過下面一個簡單的例子進行理解：

package main

import (
    "fmt"
    "time"
    "net/http"
    "context"
    "io/ioutil"
)

type Result struct{
    r *http.Response
    err error
}

func process(){
    ctx,cancel := context.WithTimeout(context.Background(),2*time.Second)
    defer cancel()
    tr := &http.Transport{}
    client := &http.Client{Transport:tr}
    c := make(chan Result,1)
    req,err := http.NewRequest("GET","http://www.google.com",nil)
    if err != nil{
        fmt.Println("http request failed,err:",err)
        return
    }
    // 若是請求成功了會將數據存入到管道中
    go func(){
        resp,err := client.Do(req)
        pack := Result{resp,err}
        c <- pack
    }()

    select{
    case <- ctx.Done():
        tr.CancelRequest(req)
        fmt.Println("timeout!")
    case res := <-c:
        defer res.r.Body.Close()
        out,_:= ioutil.ReadAll(res.r.Body)
        fmt.Printf("server response:%s",out)
    }
    return

}

func main() {
    process()
}

寫一個經過 context 保存上下文，代碼例子如：

package main

import (
    "github.com/Go-zh/net/context"
    "fmt"
)

func add(ctx context.Context,a,b int) int {
    traceId := ctx.Value("trace_id").(string)
    fmt.Printf("trace_id:%v\n",traceId)
    return a+b
}

func calc(ctx context.Context,a, b int) int{
    traceId := ctx.Value("trace_id").(string)
    fmt.Printf("trace_id:%v\n",traceId)
    //再將ctx傳入到add中
    return add(ctx,a,b)
}

func main() {
    //將ctx傳遞到calc中
    ctx := context.WithValue(context.Background(),"trace_id","123456")
    calc(ctx,20,30)

}

結合 etcd 和 context 使用

關於經過 go 鏈接 etcd 的簡單例子：(這裏有個小問題須要注意就是 etcd 的啓動方式，默認啓動可能會鏈接不上，尤爲你是在虛擬你安裝，因此須要經過以下命令啓動：
./etcd --listen-client-urls http://0.0.0.0:2371 --advertise-client-urls http://0.0.0.0:2371 --listen-peer-urls http://0.0.0.0:2381
)

package main

import (
    etcd_client "github.com/coreos/etcd/clientv3"
    "time"
    "fmt"
)

func main() {
    cli, err := etcd_client.New(etcd_client.Config{
        Endpoints:[]string{"192.168.0.118:2371"},
        DialTimeout:5*time.Second,
    })
    if err != nil{
        fmt.Println("connect failed,err:",err)
        return
    }

    fmt.Println("connect success")
    defer cli.Close()
}

下面一個例子是經過鏈接etcd，存值並取值

package main

import (
    "github.com/coreos/etcd/clientv3"
    "time"
    "fmt"
    "context"
)

func main() {
    cli,err := clientv3.New(clientv3.Config{
        Endpoints:[]string{"192.168.0.118:2371"},
        DialTimeout:5*time.Second,
    })
    if err != nil{
        fmt.Println("connect failed,err:",err)
        return
    }
    fmt.Println("connect succ")
    defer cli.Close()
    ctx,cancel := context.WithTimeout(context.Background(),time.Second)
    _,err = cli.Put(ctx,"logagent/conf/","sample_value")
    cancel()
    if err != nil{
        fmt.Println("put failed,err",err)
        return
    }
    ctx, cancel = context.WithTimeout(context.Background(),time.Second)
    resp,err := cli.Get(ctx,"logagent/conf/")
    cancel()
    if err != nil{
        fmt.Println("get failed,err:",err)
        return
    }
    for _,ev := range resp.Kvs{
        fmt.Printf("%s:%s\n",ev.Key,ev.Value)
    }
}

關於 context 官網也有一個例子很是有用，用於控制開啓的 goroutine 的退出，代碼以下：

package main

import (
    "context"
    "fmt"
)

func main() {
    // gen generates integers in a separate goroutine and
    // sends them to the returned channel.
    // The callers of gen need to cancel the context once
    // they are done consuming generated integers not to leak
    // the internal goroutine started by gen.
    gen := func(ctx context.Context) <-chan int {
        dst := make(chan int)
        n := 1
        go func() {
            for {
                select {
                case <-ctx.Done():
                    return // returning not to leak the goroutine
                case dst <- n:
                    n++
                }
            }
        }()
        return dst
    }

    ctx, cancel := context.WithCancel(context.Background())
    defer cancel() // cancel when we are finished consuming integers

    for n := range gen(ctx) {
        fmt.Println(n)
        if n == 5 {
            break
        }
    }
}

關於官網文檔中的WithDeadline演示的代碼例子：

package main

import (
    "context"
    "fmt"
    "time"
)

func main() {
    d := time.Now().Add(50 * time.Millisecond)
    ctx, cancel := context.WithDeadline(context.Background(), d)

    // Even though ctx will be expired, it is good practice to call its
    // cancelation function in any case. Failure to do so may keep the
    // context and its parent alive longer than necessary.
    defer cancel()

    select {
    case <-time.After(1 * time.Second):
        fmt.Println("overslept")
    case <-ctx.Done():
        fmt.Println(ctx.Err())
    }

}

經過上面的代碼有了一個基本的使用，那麼若是咱們經過etcd來作配置管理，若是配置更改以後，咱們如何通知對應的服務器配置更改，經過下面例子演示：

package main

import (
    "github.com/coreos/etcd/clientv3"
    "time"
    "fmt"
    "context"
)

func main() {
    cli,err := clientv3.New(clientv3.Config{
        Endpoints:[]string{"192.168.0.118:2371"},
        DialTimeout:5*time.Second,
    })
    if err != nil {
        fmt.Println("connect failed,err:",err)
        return
    }
    defer cli.Close()
    // 這裏會阻塞
    rch := cli.Watch(context.Background(),"logagent/conf/")
    for wresp := range rch{
        for _,ev := range wresp.Events{
            fmt.Printf("%s %q : %q\n", ev.Type, ev.Kv.Key, ev.Kv.Value)
        }
    }
}

實現一個 kafka 的消費者代碼的簡單例子：

package main

import (
    "github.com/Shopify/sarama"
    "strings"
    "fmt"
    "time"
)

func main() {
    consumer,err := sarama.NewConsumer(strings.Split("192.168.0.118:9092",","),nil)
    if err != nil{
        fmt.Println("failed to start consumer:",err)
        return
    }
    partitionList,err := consumer.Partitions("nginx_log")
    if err != nil {
        fmt.Println("Failed to get the list of partitions:",err)
        return
    }
    fmt.Println(partitionList)
    for partition := range partitionList{
        pc,err := consumer.ConsumePartition("nginx_log",int32(partition),sarama.OffsetNewest)
        if err != nil {
            fmt.Printf("failed to start consumer for partition %d:%s\n",partition,err)
            return
        }
        defer pc.AsyncClose()
        go func(partitionConsumer sarama.PartitionConsumer){
            for msg := range pc.Messages(){
                fmt.Printf("partition:%d Offset:%d Key:%s Value:%s",msg.Partition,msg.Offset,string(msg.Key),string(msg.Value))
            }
        }(pc)
    }
    time.Sleep(time.Hour)
    consumer.Close()

}

可是上面的代碼並非最佳代碼，由於咱們最後是經過time.sleep等待goroutine的執行，咱們能夠更改成經過sync.WaitGroup方式實現

package main

import (
    "github.com/Shopify/sarama"
    "strings"
    "fmt"
    "sync"
)

var (
    wg sync.WaitGroup
)

func main() {
    consumer,err := sarama.NewConsumer(strings.Split("192.168.0.118:9092",","),nil)
    if err != nil{
        fmt.Println("failed to start consumer:",err)
        return
    }
    partitionList,err := consumer.Partitions("nginx_log")
    if err != nil {
        fmt.Println("Failed to get the list of partitions:",err)
        return
    }
    fmt.Println(partitionList)
    for partition := range partitionList{
        pc,err := consumer.ConsumePartition("nginx_log",int32(partition),sarama.OffsetNewest)
        if err != nil {
            fmt.Printf("failed to start consumer for partition %d:%s\n",partition,err)
            return
        }
        defer pc.AsyncClose()
        go func(partitionConsumer sarama.PartitionConsumer){
            wg.Add(1)
            for msg := range partitionConsumer.Messages(){
                fmt.Printf("partition:%d Offset:%d Key:%s Value:%s",msg.Partition,msg.Offset,string(msg.Key),string(msg.Value))
            }
            wg.Done()
        }(pc)
    }

    //time.Sleep(time.Hour)
    wg.Wait()
    consumer.Close()

}

將客戶端須要收集的日誌信息放到 etcd 中

關於 etcd 處理的代碼爲：

package main

import (
    "github.com/coreos/etcd/clientv3"
    "time"
    "github.com/astaxie/beego/logs"
    "context"
    "fmt"
)

var Client *clientv3.Client
var logConfChan chan string

// 初始化etcd
func initEtcd(addr []string,keyfmt string,timeout time.Duration)(err error){

    var keys []string
    for _,ip := range ipArrays{
        //keyfmt = /logagent/%s/log_config
        keys = append(keys,fmt.Sprintf(keyfmt,ip))
    }

    logConfChan = make(chan string,10)
    logs.Debug("etcd watch key:%v timeout:%v", keys, timeout)

    Client,err = clientv3.New(clientv3.Config{
        Endpoints:addr,
        DialTimeout: timeout,
    })
    if err != nil{
        logs.Error("connect failed,err:%v",err)
        return
    }
    logs.Debug("init etcd success")
    waitGroup.Add(1)
    for _, key := range keys{
        ctx,cancel := context.WithTimeout(context.Background(),2*time.Second)
        // 從etcd中獲取要收集日誌的信息
        resp,err := Client.Get(ctx,key)
        cancel()
        if err != nil {
            logs.Warn("get key %s failed,err:%v",key,err)
            continue
        }

        for _, ev := range resp.Kvs{
            logs.Debug("%q : %q\n",  ev.Key, ev.Value)
            logConfChan <- string(ev.Value)
        }
    }
    go WatchEtcd(keys)
    return
}

func WatchEtcd(keys []string){
    // 這裏用於檢測當須要收集的日誌信息更改時及時更新
    var watchChans []clientv3.WatchChan
    for _,key := range keys{
        rch := Client.Watch(context.Background(),key)
        watchChans = append(watchChans,rch)
    }

    for {
        for _,watchC := range watchChans{
            select{
            case wresp := <-watchC:
                for _,ev:= range wresp.Events{
                    logs.Debug("%s %q : %q\n", ev.Type, ev.Kv.Key, ev.Kv.Value)
                    logConfChan <- string(ev.Kv.Value)
                }
            default:

            }
        }
        time.Sleep(time.Second)
    }
    waitGroup.Done()
}

func GetLogConf()chan string{
    return logConfChan
}

一樣的這裏增長對了限速的處理，畢竟日誌收集程序不能影響了當前業務的性能，因此增長了limit.go用於限制速度：

package main

import (
    "time"
    "sync/atomic"
    "github.com/astaxie/beego/logs"
)

type SecondLimit struct {
    unixSecond int64
    curCount int32
    limit int32
}

func NewSecondLimit(limit int32) *SecondLimit {
    secLimit := &SecondLimit{
        unixSecond:time.Now().Unix(),
        curCount:0,
        limit:limit,
    }
    return secLimit
}

func (s *SecondLimit) Add(count int) {
    sec := time.Now().Unix()
    if sec == s.unixSecond {
        atomic.AddInt32(&s.curCount,int32(count))
        return
    }
    atomic.StoreInt64(&s.unixSecond,sec)
    atomic.StoreInt32(&s.curCount, int32(count))
}

func (s *SecondLimit) Wait()bool {
    for {
        sec := time.Now().Unix()
        if (sec == atomic.LoadInt64(&s.unixSecond)) && s.curCount == s.limit {
            time.Sleep(time.Microsecond)
            logs.Debug("limit is running,limit:%d s.curCount:%d",s.limit,s.curCount)
            continue
        }

        if sec != atomic.LoadInt64(&s.unixSecond) {
            atomic.StoreInt64(&s.unixSecond,sec)
            atomic.StoreInt32(&s.curCount,0)
        }
        logs.Debug("limit is exited")
        return false
    }
}

小結

此次基本實現了日誌收集的前半段的處理，後面將把日誌扔到es中，並最終在頁面上呈現
做者;coder
原文連接：https://www.cnblogs.com/zhaof/p/8910761.html

相關文章
相關標籤/搜索
每日一句
    每一个你不满意的现在，都有一个你没有努力的曾经。
本站公眾號
   歡迎關注本站公眾號,獲取更多信息
聯繫我們 最近搜索 最新文章 沪ICP备13005482号-10 MyBatis教程 SQL 教程 MySQL教程 Java 教程 Thymeleaf 教程 Hibernate教程 Spring教程 Redis教程