[Golang] kafka集羣搭建和golang版生產者和消費者

1、kafka集羣搭建　　

　　至於kafka是什麼我都很少作介紹了，網上寫的已經很是詳盡了。

 (沒安裝java環境的須要先安裝 yum -y install java-1.8.0-openjdk*)

1. 下載zookeeper  https://zookeeper.apache.org/releases.html

2. 下載kafka http://kafka.apache.org/downloads

3. 啓動zookeeper集羣（個人示例是3臺機器，後面的kafka也同樣，這裏就以1臺代指3臺，固然你也能夠只開1臺）

　　1)配置zookeeper。 修改複製一份 zookeeper-3.4.13/conf/zoo_sample.cfg 更名成zoo.cfg。修改如下幾個參數，改爲適合本身機器的。

　　dataDir=/home/test/zookeeper/data
　　dataLogDir=/home/test/zookeeper/log
　　server.1=10.22.1.1:2888:3888
　　server.2=10.22.1.2:2888:3888
　　server.3=10.22.1.3:2888:3888

　　2) 建立myid文件，肯定機器編號。分別在3臺機器的/home/test/zookeeper/data目錄執行分別執行命令 echo 1 > myid(注意ip爲10.22.1.2把1改爲2，見上面的配置)

　　3) 啓動zookeeper集羣。分別進入目錄zookeeper-3.4.13/bin 執行 sh zkServer.sh start

4. 啓動kafka集羣

　　1) 配置kafka。進入kafka_2.11-2.2.0/config。複製3份，分別爲server1.properties，server2.properties，server3.properties。修改如下幾項(注意對應的機器id)

log.dirs和zookeeper.connect 是同樣的。broker.id和listeners分別填對應的id和ip

broker.id=1
listeners=PLAINTEXT://10.22.1.1:9092
log.dirs=/home/test/kafka/log
zookeeper.connect=10.22.1.1:2181,10.22.1.2:2181,10.22.1.3:2181

　　2) 啓動kafka集羣。分別進入kafka_2.11-2.2.0/bin目錄,分別執行sh kafka-server-start.sh ../config/server1.properties (第2臺用server2.properties配置文件)

　

2、Golang生產者和消費者

　　目前比較流行的golang版的kafka客戶端庫有兩個：

　　1. https://github.com/Shopify/sarama

　　2. https://github.com/confluentinc/confluent-kafka-go

　　至於誰好誰壞本身去分辨，我用的是第1個，star比較多的。

 

1. kafka生產者代碼

　　這裏有2點要說明：

　　1)  config.Producer.Partitioner = sarama.NewRandomPartitioner，我分partition用的是隨機，若是你想穩定分paritition的話能夠自定義，還有輪詢和hash方式

　　2) 個人topic是走的外部配置，能夠根據本身的需求修改

// Package kafka_producer kafka 生產者的包裝
package kafka_producer

import (
    "github.com/Shopify/sarama"
    "strings"
    "sync"
    "time"

    "github.com/alecthomas/log4go"
)

// Config 配置
type Config struct {
    Topic      string `xml:"topic"`
    Broker     string `xml:"broker"`
    Frequency  int    `xml:"frequency"`
    MaxMessage int    `xml:"max_message"`
}

type Producer struct {
    producer sarama.AsyncProducer

    topic     string
    msgQ      chan *sarama.ProducerMessage
    wg        sync.WaitGroup
    closeChan chan struct{}
}

// NewProducer 構造KafkaProducer
func NewProducer(cfg *Config) (*Producer, error) {

    config := sarama.NewConfig()
    config.Producer.RequiredAcks = sarama.NoResponse                                  // Only wait for the leader to ack
    config.Producer.Compression = sarama.CompressionSnappy                            // Compress messages
    config.Producer.Flush.Frequency = time.Duration(cfg.Frequency) * time.Millisecond // Flush batches every 500ms
    config.Producer.Partitioner = sarama.NewRandomPartitioner

    p, err := sarama.NewAsyncProducer(strings.Split(cfg.Broker, ","), config)
    if err != nil {
        return nil, err
    }
    ret := &Producer{
        producer:  p,
        topic:     cfg.Topic,
        msgQ:      make(chan *sarama.ProducerMessage, cfg.MaxMessage),
        closeChan: make(chan struct{}),
    }

    return ret, nil
}

// Run 運行
func (p *Producer) Run() {

    p.wg.Add(1)
    go func() {
        defer p.wg.Done()

    LOOP:
        for {
            select {
            case m := <-p.msgQ:
                p.producer.Input() <- m
            case err := <-p.producer.Errors():
                if nil != err && nil != err.Msg {
                    l4g.Error("[producer] err=[%s] topic=[%s] key=[%s] val=[%s]", err.Error(), err.Msg.Topic, err.Msg.Key, err.Msg.Value)
                }
            case <-p.closeChan:
                break LOOP
            }

        }
    }()

    for hasTask := true; hasTask; {
        select {
        case m := <-p.msgQ:
            p.producer.Input() <- m
        default:
            hasTask = false
        }
    }

}

// Close 關閉
func (p *Producer) Close() error {
    close(p.closeChan)
    l4g.Warn("[producer] is quiting")
    p.wg.Wait()
    l4g.Warn("[producer] quit over")

    return p.producer.Close()
}

// Log 發送log
func (p *Producer) Log(key string, val string) {
    msg := &sarama.ProducerMessage{
        Topic: p.topic,
        Key:   sarama.StringEncoder(key),
        Value: sarama.StringEncoder(val),
    }

    select {
    case p.msgQ <- msg:
        return
    default:
        l4g.Error("[producer] err=[msgQ is full] key=[%s] val=[%s]", msg.Key, msg.Value)
    }
}

 

2. kafka消費者

　　幾點說明：

　　1) kafka必定要選用支持集羣的版本

　　2) 裏面帶了建立topic，刪除topic，打印topic的工具

　　3) replication是外面配置的

　　4) 開多個consumer須要在建立topic時設置多個partition。官方的示例當開多個consumer的時候會崩潰，我這個版本不會，我給官方提交了一個PR，還不知道有沒有采用

 

// Package main Kafka消費者
package main

import (
    "context"
    "encoding/xml"
    "flag"
    "fmt"
    "io/ioutil"
    "log"
    "os"
    "os/signal"
    "runtime"
    "strings"
    "syscall"
    "time"

    "github.com/Shopify/sarama"
    "github.com/alecthomas/log4go"
)

// Consumer Consumer配置
type ConsumerConfig struct {
    Topic       []string `xml:"topic"`
    Broker      string   `xml:"broker"`
    Partition   int32    `xml:"partition"`
    Replication int16    `xml:"replication"`
    Group       string   `xml:"group"`
    Version     string   `xml:"version"`
}

var (
    configFile = "" // 配置路徑
    initTopic  = false
    listTopic  = false
    delTopic   = ""
    cfg        = &Config{}
)

// Config 配置
type Config struct {
    Consumer ConsumerConfig `xml:"consumer"`
}

func init() {
    flag.StringVar(&configFile, "config", "../config/consumer.xml", "config file ")
    flag.BoolVar(&initTopic, "init", initTopic, "create topic")
    flag.BoolVar(&listTopic, "list", listTopic, "list topic")
    flag.StringVar(&delTopic, "del", delTopic, "delete topic")

}

func main() {

    runtime.GOMAXPROCS(runtime.NumCPU())

    defer func() {
        time.Sleep(time.Second)
        log4go.Warn("[main] consumer quit over!")
        log4go.Global.Close()
    }()

    contents, _ := ioutil.ReadFile(configFile)
    xml.Unmarshal(contents, cfg)

    // sarama的logger
    sarama.Logger = log.New(os.Stdout, fmt.Sprintf("[%s]", "consumer"), log.LstdFlags)

    // 指定kafka版本，必定要支持kafka集羣
    version, err := sarama.ParseKafkaVersion(cfg.Consumer.Version)
    if err != nil {
        panic(err)
    }
    config := sarama.NewConfig()
    config.Version = version
    config.Consumer.Offsets.Initial = sarama.OffsetOldest

    // 工具
    if tool(cfg, config) {
        return
    }

    // kafka consumer client
    ctx, cancel := context.WithCancel(context.Background())
    client, err := sarama.NewConsumerGroup(strings.Split(cfg.Consumer.Broker, ","), cfg.Consumer.Group, config)
    if err != nil {
        panic(err)
    }

    consumer := Consumer{}
    go func() {
        for {
            err := client.Consume(ctx, cfg.Consumer.Topic, &consumer)
            if err != nil {
                log4go.Error("[main] client.Consume error=[%s]", err.Error())
                // 5秒後重試
                time.Sleep(time.Second * 5)
            }
        }
    }()

    // os signal
    sigterm := make(chan os.Signal, 1)
    signal.Notify(sigterm, syscall.SIGINT, syscall.SIGTERM)

    <-sigterm

    cancel()
    err = client.Close()
    if err != nil {
        panic(err)
    }

    log4go.Info("[main] consumer is quiting")
}

func tool(cfg *Config, config *sarama.Config) bool {
    if initTopic || listTopic || len(delTopic) > 0 {
        ca, err := sarama.NewClusterAdmin(strings.Split(cfg.Consumer.Broker, ","), config)
        if nil != err {
            panic(err)
        }

        if len(delTopic) > 0 { // 刪除Topic
            if err := ca.DeleteTopic(delTopic); nil != err {
                panic(err)
            }
            log4go.Info("delete ok topic=[%s]\n", delTopic)
        } else if initTopic { // 初始化Topic
            if detail, err := ca.ListTopics(); nil != err {
                panic(err)
            } else {
                for _, v := range cfg.Consumer.Topic {
                    if d, ok := detail[v]; ok {
                        if cfg.Consumer.Partition > d.NumPartitions {
                            if err := ca.CreatePartitions(v, cfg.Consumer.Partition, nil, false); nil != err {
                                panic(err)
                            }
                            log4go.Info("alter topic ok", v, cfg.Consumer.Partition)
                        }

                    } else {
                        if err := ca.CreateTopic(v, &sarama.TopicDetail{NumPartitions: cfg.Consumer.Partition, ReplicationFactor: cfg.Consumer.Replication}, false); nil != err {
                            panic(err)
                        }
                        log4go.Info("create topic ok", v)
                    }
                }
            }
        }

        // 顯示Topic列表
        if detail, err := ca.ListTopics(); nil != err {
            log4go.Info("ListTopics error", err)
        } else {
            for k := range detail {
                log4go.Info("[%s] %+v", k, detail[k])
            }
        }

        if err := ca.Close(); nil != err {
            panic(err)
        }

        return true
    }
    return false
}

type Consumer struct {
}

func (consumer *Consumer) Setup(s sarama.ConsumerGroupSession) error {
    return nil
}

func (consumer *Consumer) Cleanup(s sarama.ConsumerGroupSession) error {
    return nil
}

func (consumer *Consumer) ConsumeClaim(session sarama.ConsumerGroupSession, claim sarama.ConsumerGroupClaim) error {
    for message := range claim.Messages() {
        key := string(message.Key)
        val := string(message.Value)
        log4go.Info("%s-%s", key, val)
        session.MarkMessage(message, "")
    }

    return nil
}