8-kafka消費者0.8 0.9 版本高級API和簡單API

使用Old Consumer High Level API編寫consumer

第一步：編寫具體處理消息的類

import java.io.UnsupportedEncodingException;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import kafka.consumer.ConsumerIterator;
import kafka.consumer.KafkaStream;
import kafka.message.MessageAndMetadata;


public class Consumerwork implements Runnable {
    private static Logger LOG = LoggerFactory.getLogger(Consumerwork.class);
     @SuppressWarnings("rawtypes")
    private KafkaStream m_stream;
     private int m_threadNumber;
     @SuppressWarnings("rawtypes")
    public Consumerwork(KafkaStream a_stream,int a_threadNumber) {
        // TODO Auto-generated constructor stub
         m_threadNumber = a_threadNumber;
         m_stream = a_stream;
    }

    @SuppressWarnings("unchecked")
    @Override
    public void run() {
        // TODO Auto-generated method stub
        ConsumerIterator<byte[], byte[]> it = m_stream.iterator();
           while (it.hasNext())
                try {
                    MessageAndMetadata<byte[], byte[]> thisMetadata=it.next();
                    String jsonStr = new String(thisMetadata.message(),"utf-8") ;
                    LOG.info("Thread " + m_threadNumber + ": " +jsonStr);
                    LOG.info("partion"+thisMetadata.partition()+",offset:"+thisMetadata.offset());
                    try {
                        Thread.sleep(1000);
                    } catch (InterruptedException e) {
                        // TODO Auto-generated catch block
                        e.printStackTrace();
                    }
                } catch (UnsupportedEncodingException e) {
                    // TODO Auto-generated catch block
                    e.printStackTrace();
                }
    }
}

第二步：編寫啓動Consumer主類

import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.Properties;
import java.util.Scanner;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.TimeUnit;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import kafka.consumer.ConsumerConfig;
import kafka.consumer.KafkaStream;
import kafka.javaapi.consumer.ConsumerConnector;
public class ConsumerGroup {
    private final ConsumerConnector consumer;
    private final String topic;
    private ExecutorService executor;
    private static Logger LOG = LoggerFactory.getLogger(ConsumerGroup.class);
    public ConsumerGroup(String a_zookeeper, String a_groupId, String a_topic) {
        consumer = kafka.consumer.Consumer.createJavaConsumerConnector(createConsumerConfig(a_zookeeper, a_groupId));
        this.topic = a_topic;
    }
    public static void main(String[] args) {
        Scanner sc = new Scanner(System.in);
        System.out.println("請輸入zookeeper集羣地址(如zk1:2181,zk2:2181,zk3:2181)：");
        String zooKeeper = sc.nextLine(); 
        System.out.println("請輸入指定的消費group名稱：");
        String groupId = sc.nextLine(); 
        System.out.println("請輸入指定的消費topic名稱：");
        String topic = sc.nextLine(); 
        System.out.println("請輸入指定的消費處理線程數：");
        int threads = sc.nextInt();
        LOG.info("Starting consumer kafka messages with zk:" + zooKeeper + " and the topic is " + topic);
        ConsumerGroup example = new ConsumerGroup(zooKeeper, groupId, topic);
        example.run(threads);

        try {
            Thread.sleep(1000);
        } catch (InterruptedException ie) {

        }
        // example.shutdown();
    }

    private void shutdown() {
        // TODO Auto-generated method stub
        if (consumer != null)
            consumer.shutdown();
        if (executor != null)
            executor.shutdown();
        try {
            if (!executor.awaitTermination(5000, TimeUnit.MILLISECONDS)) {
                LOG.info("Timed out waiting for consumer threads to shut down, exiting uncleanly");
            }
        } catch (InterruptedException e) {
            LOG.info("Interrupted during shutdown, exiting uncleanly");
        }
    }

    private void run(int a_numThreads) {
        // TODO Auto-generated method stub
        Map<String, Integer> topicCountMap = new HashMap<String, Integer>();
        topicCountMap.put(topic, new Integer(a_numThreads));
        Map<String, List<KafkaStream<byte[], byte[]>>> consumerMap = consumer.createMessageStreams(topicCountMap);
        List<KafkaStream<byte[], byte[]>> streams = consumerMap.get(topic);

        // now launch all the threads
        //
        executor = Executors.newFixedThreadPool(a_numThreads);

        // now create an object to consume the messages
        //
        int threadNumber = 0;
        LOG.info("the streams size is "+streams.size());
        for (final KafkaStream stream : streams) {
            executor.submit(new com.goodix.kafka.oldconsumer.Consumerwork(stream, threadNumber));
    //      consumer.commitOffsets();
            threadNumber++;
        }

    }

    private ConsumerConfig createConsumerConfig(String a_zookeeper, String a_groupId) {
        // TODO Auto-generated method stub
        Properties props = new Properties();
        props.put("zookeeper.connect", a_zookeeper);
        props.put("group.id", a_groupId);
        props.put("zookeeper.session.timeout.ms", "60000");
        props.put("zookeeper.sync.time.ms", "200");
        props.put("auto.commit.interval.ms", "1000");
        props.put("auto.offset.reset", "smallest");
//      props.put("rebalance.max.retries", "5");
//      props.put("rebalance.backoff.ms", "15000");
        return new ConsumerConfig(props);
    }

}

1. topicCountMap.put(topic, new Integer(a_numThreads)) 是告訴Kafka我有多少個線程來處理消息。

(1). 這個線程數必須是小等於topic的partition分區數；能夠經過./kafka-topics.sh --describe --zookeeper "172.16.49.173:2181" --topic "producer_test"命令來查看分區的狀況 
(2). kafka會根據partition.assignment.strategy指定的分配策略來指定線程消費那些分區的消息；這裏沒有單獨配置該項便是採用的默認值range策略（按照階段平均分配）。好比分區有10個、線程數有3個，則線程 1消費0,1,2,3，線程2消費4,5,6,線程3消費7,8,9。另一種是roundrobin(循環分配策略)，官方文檔中寫有使用該策略有兩個前提條件的，因此通常不要去設定。 
(3). 通過測試：consumerMap.get(topic).size()，應該是得到的目前該topic有數據的分區數 
(4). stream即指的是來自一個或多個服務器上的一個或者多個partition的消息。每個stream都對應一個單線程處理。所以，client可以設置知足本身需求的stream數目。總之，一個stream也許表明了多個服務器partion的消息的聚合，可是每個 partition都只能到一個stream

2. Executors.newFixedThreadPool(a_numThreads)是建立一個建立固定容量大小的緩衝池：每次提交一個任務就建立一個線程，直到線程達到線程池的最大大小。線程池的大小一旦達到最大值就會保持不變，若是某個線程由於執行異常而結束，那麼線程池會補充一個新線程。

3. props.put(「auto.offset.reset」, 「smallest」) 是指定從最小沒有被消費offset開始；若是沒有指定該項則是默認的爲largest，這樣的話該consumer就得不到生產者先產生的消息。

4. 要使用old consumer API須要引用kafka_2.11以及kafka-clients。

<dependency>
<groupId>org.apache.kafka</groupId>
<artifactId>kafka_2.11</artifactId>
<version>0.10.0.0</version>
</dependency>
<dependency>
<groupId>org.apache.kafka</groupId>
<artifactId>kafka-clients</artifactId>
<version>0.10.0.0</version>
</dependency>

 

使用Old SimpleConsumerAPI編寫consumer

使用的場景

這是一個更加底層和複雜的API，因爲使用該API須要本身控制的項比較多，也比較複雜，官方給出了一些合適的適用場景，也能夠理解成爲這些場景是High Level Consumer API 不可以作到的

1. 針對一個消息讀取屢次

2. 在一個process中，僅僅處理一個topic中的一個partitions

3. 使用事務，確保每一個消息只被處理一次

須要處理的事情

1. 必須在程序中跟蹤offset值

2. 必須找出指定Topic Partition中的lead broker

3. 必須處理broker的變更

使用SimpleConsumer的步驟

首先，你必須知道讀哪一個topic的哪一個partition 而後，找到負責該partition的broker leader，從而找到存有該partition副本的那個broker

再者，本身去寫request並fetch數據 最終，還要注意須要識別和處理broker leader的改變

package com.goodix.kafka.oldconsumer;
import kafka.api.FetchRequest;
import kafka.api.FetchRequestBuilder;
import kafka.api.PartitionOffsetRequestInfo;
import kafka.common.ErrorMapping;
import kafka.common.TopicAndPartition;
import kafka.javaapi.*;
import kafka.javaapi.consumer.SimpleConsumer;
import kafka.message.MessageAndOffset;

import java.nio.ByteBuffer;
import java.util.ArrayList;
import java.util.Collections;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.Scanner;

import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

public class SimpleExample {
    private static Logger LOG = LoggerFactory.getLogger(SimpleExample.class);
    public static void main(String args[]) {
        SimpleExample example = new SimpleExample();
        Scanner sc = new Scanner(System.in);
        System.out.println("請輸入broker節點的ip地址(如172.16.49.173)");
        String brokerIp = sc.nextLine(); 
        List<String> seeds = new ArrayList<String>();
        seeds.add(brokerIp);
        System.out.println("請輸入broker節點端口號(如9092)");
        int port = Integer.parseInt( sc.nextLine());
        System.out.println("請輸入要訂閱的topic名稱(如test)");
        String topic = sc.nextLine();
        System.out.println("請輸入要訂閱要查找的分區(如0)");
        int partition = Integer.parseInt( sc.nextLine());
        System.out.println("請輸入最大讀取消息數量(如10000)");
        long maxReads = Long.parseLong( sc.nextLine());

        try {
            example.run(maxReads, topic, partition, seeds, port);
        } catch (Exception e) {
            LOG.error("Oops:" + e);
             e.printStackTrace();
        }
    }

    private List<String> m_replicaBrokers = new ArrayList<String>();

    public SimpleExample() {
        m_replicaBrokers = new ArrayList<String>();
    }

    public void run(long a_maxReads, String a_topic, int a_partition, List<String> a_seedBrokers, int a_port) throws Exception {
        // find the meta data about the topic and partition we are interested in
        //獲取指定Topic partition的元數據  
        PartitionMetadata metadata = findLeader(a_seedBrokers, a_port, a_topic, a_partition);
        if (metadata == null) {
            LOG.error("Can't find metadata for Topic and Partition. Exiting");
            return;
        }
        if (metadata.leader() == null) {
            LOG.error("Can't find Leader for Topic and Partition. Exiting");
            return;
        }
        String leadBroker = metadata.leader().host();
        String clientName = "Client_" + a_topic + "_" + a_partition;
        SimpleConsumer consumer = new SimpleConsumer(leadBroker, a_port, 100000, 64 * 1024, clientName);
        long readOffset = getLastOffset(consumer,a_topic, a_partition, kafka.api.OffsetRequest.EarliestTime(), clientName);

        int numErrors = 0;
        while (a_maxReads > 0) {
            if (consumer == null) {
                consumer = new SimpleConsumer(leadBroker, a_port, 100000, 64 * 1024, clientName);
            }
            FetchRequest req = new FetchRequestBuilder()
                    .clientId(clientName)
                    .addFetch(a_topic, a_partition, readOffset, 100000) // Note: this fetchSize of 100000 might need to be increased if large batches are written to Kafka
                    .build();
            FetchResponse fetchResponse = consumer.fetch(req);

            if (fetchResponse.hasError()) {
                numErrors++;
                // Something went wrong!
                short code = fetchResponse.errorCode(a_topic, a_partition);
                LOG.error("Error fetching data from the Broker:" + leadBroker + " Reason: " + code);
                if (numErrors > 5) break;
                if (code == ErrorMapping.OffsetOutOfRangeCode())  {
                    // We asked for an invalid offset. For simple case ask for the last element to reset
                    readOffset = getLastOffset(consumer,a_topic, a_partition, kafka.api.OffsetRequest.LatestTime(), clientName);
                    continue;
                }
                consumer.close();
                consumer = null;
                leadBroker = findNewLeader(leadBroker, a_topic, a_partition, a_port);
                continue;
            }
            numErrors = 0;

            long numRead = 0;
            for (MessageAndOffset messageAndOffset : fetchResponse.messageSet(a_topic, a_partition)) {
                long currentOffset = messageAndOffset.offset();
                if (currentOffset < readOffset) {
                    LOG.error("Found an old offset: " + currentOffset + " Expecting: " + readOffset);
                    continue;
                }
                readOffset = messageAndOffset.nextOffset();
                ByteBuffer payload = messageAndOffset.message().payload();

                byte[] bytes = new byte[payload.limit()];
                payload.get(bytes);
                LOG.info("the messag's offset is :"+String.valueOf(messageAndOffset.offset()) + " and the value is :" + new String(bytes, "UTF-8"));
                numRead++;
                a_maxReads--;
            }

            if (numRead == 0) {
                try {
                    Thread.sleep(1000);
                } catch (InterruptedException ie) {
                }
            }
        }
        if (consumer != null) consumer.close();
    }

    public static long getLastOffset(SimpleConsumer consumer, String topic, int partition,
                                     long whichTime, String clientName) {
        TopicAndPartition topicAndPartition = new TopicAndPartition(topic, partition);
        Map<TopicAndPartition, PartitionOffsetRequestInfo> requestInfo = new HashMap<TopicAndPartition, PartitionOffsetRequestInfo>();
        requestInfo.put(topicAndPartition, new PartitionOffsetRequestInfo(whichTime, 1));
        kafka.javaapi.OffsetRequest request = new kafka.javaapi.OffsetRequest(
                requestInfo, kafka.api.OffsetRequest.CurrentVersion(), clientName);
        OffsetResponse response = consumer.getOffsetsBefore(request);

        if (response.hasError()) {
            LOG.error("Error fetching data Offset Data the Broker. Reason: " + response.errorCode(topic, partition) );
            return 0;
        }
        long[] offsets = response.offsets(topic, partition);
        return offsets[0];
    }
    /**
     * 找一個leader broker
     * 遍歷每一個broker，取出該topic的metadata，而後再遍歷其中的每一個partition metadata，若是找到咱們要找的partition就返回
     * 根據返回的PartitionMetadata.leader().host()找到leader broker
     * @param a_oldLeader
     * @param a_topic
     * @param a_partition
     * @param a_port
     * @return
     * @throws Exception
     */
    private String findNewLeader(String a_oldLeader, String a_topic, int a_partition, int a_port) throws Exception {
        for (int i = 0; i < 3; i++) {
            boolean goToSleep = false;
            PartitionMetadata metadata = findLeader(m_replicaBrokers, a_port, a_topic, a_partition);
            if (metadata == null) {
                goToSleep = true;
            } else if (metadata.leader() == null) {
                goToSleep = true;
            } else if (a_oldLeader.equalsIgnoreCase(metadata.leader().host()) && i == 0) {
                // first time through if the leader hasn't changed give ZooKeeper a second to recover
                // second time, assume the broker did recover before failover, or it was a non-Broker issue
                //
                goToSleep = true;
            } else {
                return metadata.leader().host();
            }
            if (goToSleep) {
                try {
                    Thread.sleep(1000);
                } catch (InterruptedException ie) {
                }
            }
        }
        LOG.error("Unable to find new leader after Broker failure. Exiting");
        throw new Exception("Unable to find new leader after Broker failure. Exiting");
    }
    /**
     * 
     * @param a_seedBrokers
     * @param a_port
     * @param a_topic
     * @param a_partition
     * @return
     */
    private PartitionMetadata findLeader(List<String> a_seedBrokers, int a_port, String a_topic, int a_partition) {
        PartitionMetadata returnMetaData = null;

        loop:
        for (String seed : a_seedBrokers) { //遍歷每一個broker 
            SimpleConsumer consumer = null;
            try {
                // 建立Simple Consumer，
                consumer = new SimpleConsumer(seed, a_port, 100000, 64 * 1024, "leaderLookup");
                List<String> topics = Collections.singletonList(a_topic);
                TopicMetadataRequest req = new TopicMetadataRequest(topics);
                //發送TopicMetadata Request請求
                kafka.javaapi.TopicMetadataResponse resp = consumer.send(req);
                //取到Topic的Metadata 
                List<TopicMetadata> metaData = resp.topicsMetadata();
                //遍歷每一個partition的metadata
                for (TopicMetadata item : metaData) {
                    for (PartitionMetadata part : item.partitionsMetadata()) {
                        // 判斷是不是要找的partition
                        if (part.partitionId() == a_partition) {
                            returnMetaData = part;
                            //找到就返回
                            break loop;
                        }
                    }
                }
            } catch (Exception e) {
                LOG.info("Error communicating with Broker [" + seed + "] to find Leader for [" + a_topic
                        + ", " + a_partition + "] Reason: " + e);
            } finally {
                if (consumer != null) consumer.close();
            }
        }
        if (returnMetaData != null) {
            m_replicaBrokers.clear();
            for (kafka.cluster.BrokerEndPoint replica : returnMetaData.replicas()) {
                m_replicaBrokers.add(replica.host());
            }
        }
        return returnMetaData;
    }
}