最近测试跟我说，某个应用消费不到交易的消息。登录到Kafka Broker看下了下日志，发现一直在报错：

java.lang.IllegalArgumentException: Magic v0 does not support record headers
    at org.apache.kafka.common.record.MemoryRecordsBuilder.appendWithOffset(MemoryRecordsBuilder.java:385)
    at org.apache.kafka.common.record.MemoryRecordsBuilder.append(MemoryRecordsBuilder.java:568)
    at org.apache.kafka.common.record.AbstractRecords.convertRecordBatch(AbstractRecords.java:117)
    at org.apache.kafka.common.record.AbstractRecords.downConvert(AbstractRecords.java:98)
    at org.apache.kafka.common.record.FileRecords.downConvert(FileRecords.java:245)
    at kafka.server.KafkaApis$$anonfun$kafka$server$KafkaApis$$convertedPartitionData$1$1$$anonfun$apply$5.apply(KafkaApis.scala:523)
    at kafka.server.KafkaApis$$anonfun$kafka$server$KafkaApis$$convertedPartitionData$1$1$$anonfun$apply$5.apply(KafkaApis.scala:521)
    at scala.Option.map(Option.scala:146)
    at kafka.server.KafkaApis$$anonfun$kafka$server$KafkaApis$$convertedPartitionData$1$1.apply(KafkaApis.scala:521)
    at kafka.server.KafkaApis$$anonfun$kafka$server$KafkaApis$$convertedPartitionData$1$1.apply(KafkaApis.scala:511)
    at scala.Option.flatMap(Option.scala:171)
    at kafka.server.KafkaApis.kafka$server$KafkaApis$$convertedPartitionData$1(KafkaApis.scala:511)
    at kafka.server.KafkaApis$$anonfun$kafka$server$KafkaApis$$createResponse$2$1.apply(KafkaApis.scala:559)
    at kafka.server.KafkaApis$$anonfun$kafka$server$KafkaApis$$createResponse$2$1.apply(KafkaApis.scala:558)
    at scala.collection.Iterator$class.foreach(Iterator.scala:891)
    at scala.collection.AbstractIterator.foreach(Iterator.scala:1334)
    at scala.collection.IterableLike$class.foreach(IterableLike.scala:72)
    at scala.collection.AbstractIterable.foreach(Iterable.scala:54)
    at kafka.server.KafkaApis.kafka$server$KafkaApis$$createResponse$2(KafkaApis.scala:558)
    at kafka.server.KafkaApis$$anonfun$kafka$server$KafkaApis$$fetchResponseCallback$1$1.apply$mcVI$sp(KafkaApis.scala:579)
    at kafka.server.ClientQuotaManager.recordAndThrottleOnQuotaViolation(ClientQuotaManager.scala:196)
    at kafka.server.KafkaApis.sendResponseMaybeThrottle(KafkaApis.scala:2012)
    at kafka.server.KafkaApis.kafka$server$KafkaApis$$fetchResponseCallback$1(KafkaApis.scala:578)
    at kafka.server.KafkaApis$$anonfun$kafka$server$KafkaApis$$processResponseCallback$1$1.apply$mcVI$sp(KafkaApis.scala:598)
    at kafka.server.ClientQuotaManager.recordAndThrottleOnQuotaViolation(ClientQuotaManager.scala:196)
    at kafka.server.ClientQuotaManager.recordAndMaybeThrottle(ClientQuotaManager.scala:188)
    at kafka.server.KafkaApis.kafka$server$KafkaApis$$processResponseCallback$1(KafkaApis.scala:597)
    at kafka.server.KafkaApis$$anonfun$handleFetchRequest$1.apply(KafkaApis.scala:614)
    at kafka.server.KafkaApis$$anonfun$handleFetchRequest$1.apply(KafkaApis.scala:614)
    at kafka.server.ReplicaManager.fetchMessages(ReplicaManager.scala:639)
    at kafka.server.KafkaApis.handleFetchRequest(KafkaApis.scala:606)
    at kafka.server.KafkaApis.handle(KafkaApis.scala:98)
    at kafka.server.KafkaRequestHandler.run(KafkaRequestHandler.scala:66)
    at java.lang.Thread.run(Thread.java:748)

问了下相关开发人员，发现最近有个版本需要在Kafka信息的Headers中增加LogId来做交易跟踪，结合错误信息中提示消费者Api版本太低，不支持header信息，导致出错，让开发人员去掉header后，消费者可以正常消费消息

下面我们来模拟重现

Kafka版本：0.11.0

生产者代码:

​ 写了个拦截器，为每条消息的header中添加LOG_ID

public  class KafkaProducerInterceptor<K, V> implements ProducerInterceptor<K, V> {

    @Override
    public void configure(Map<String, ?> configs) {}


    @Override
    public void onAcknowledgement(RecordMetadata metadata, Exception exception) {}

    @Override
    public void close() {}

    @Override
    public ProducerRecord onSend(ProducerRecord<K, V> record) {
        String uuid = UUID.randomUUID().toString();
        record.headers().add("LOG_ID",uuid.getBytes());
        return record;
    }

}

public class App 
{
    public static void main( String[] args ) throws InterruptedException
    {
        Properties props = new Properties();
        // broker地址
        props.put("bootstrap.servers", "localhost:9092");

        // 请求时候需要验证
        props.put("acks", "all");

        // 请求失败时候需要重试
        props.put("retries", 1);

        // 内存缓存区大小
        props.put("buffer.memory", 33554432);

        // 指定消息key序列化方式
        props.put("key.serializer", "org.apache.kafka.common.serialization.StringSerializer");

        // 指定消息本身的序列化方式
        props.put("value.serializer", "org.apache.kafka.common.serialization.StringSerializer");

        
        List<String> interceptors = new ArrayList<>();
        interceptors.add("com.df.KafkaTest.KafkaProducerInterceptor"); 
        props.put(ProducerConfig.INTERCEPTOR_CLASSES_CONFIG, interceptors);
        Producer<String, String> producer = new KafkaProducer<>(props);
        
        for (int i = 10; i < 20; i++)
            try {
                
                AccessInfo ai = new AccessInfo();
                ai.setAccessId("123456");
                ai.setAccessName("源码婆媳"+Integer.toString(i));
                ai.setBusScope("01");
                ai.setIconUrl("http://www.baidu.com");
                ProducerRecord record = new ProducerRecord<>("testTopic",0,"H"+Integer.toString(i),JSON.toJSONString(ai));
                producer.send(record).get();
            } catch (ExecutionException e) {
                e.printStackTrace();
            }

        System.out.println("Message sent successfully");
        producer.close();
    }
}

消费者代码

使用高版本Api的客户端：

public class KafkaConsumerAsync {
 
    public static void main(String[] args) throws InterruptedException {
 
        // 1、准备配置文件
        String kafkas = "127.0.0.1:9092";
        Properties props = new Properties();
        //kafka连接信息
        props.put("bootstrap.servers",kafkas);
        //消费者组id
        props.put("group.id", "testTopic-002");
        //是否自动提交offset
        props.put("enable.auto.commit", "false");
        //在没有offset的情况下采取的拉取策略
        props.put("auto.offset.reset", "earliest");
        //自动提交时间间隔
        props.put("auto.commit.interval.ms", "1000");
        //设置一次fetch请求取得的数据最大为1k
        props.put("fetch.max.bytes", "1024");
        //key反序列化
        props.put("key.deserializer", "org.apache.kafka.common.serialization.StringDeserializer");
        //value反序列化
        props.put("value.deserializer", "org.apache.kafka.common.serialization.StringDeserializer");
 
        String topic = "testTopic";
        // 2、创建KafkaConsumer
        KafkaConsumer<String, String> consumer = new KafkaConsumer<>(props);
        // 3、订阅数据，不给定监听器
        consumer.subscribe(Collections.singleton(topic));
 
        try{
            //最少处理100条
            int minCommitSize = 100;
            //定义计数器
            int icount = 0;
            // 4、获取数据
            while (true) {
                ConsumerRecords<String, String> records = consumer.poll(100);
                for (ConsumerRecord<String, String> record : records) {
                    System.out.printf("topic = %s,partition = %d,offset = %d, key = %s, value = %s%n", record.topic(), record.partition(),record.offset(), record.key(), record.value());
                    icount++;
                }
            
                //在业务逻辑处理成功后提交offset
                if(icount >= minCommitSize){
                    //满足最少消费100条,再进行异步提交
                    consumer.commitAsync(new OffsetCommitCallback() {
                        @Override public void onComplete(Map<TopicPartition, OffsetAndMetadata> offsets, Exception exception) {
                            if(exception == null){
                                System.out.println("commit success");
                            }else {
                                //提交失败，对应处理
                                System.out.println("commit failed");
                            }
                        }
                    });
                    
                    //计数器归零
                    icount = 0 ;
                }
            }
        }catch (Exception e){
            e.printStackTrace();
        }finally {
            //关闭连接
            consumer.close();
        }
    }
}

执行结果：

[28 10:20:08,923 INFO ] [main] internals.AbstractCoordinator - [Consumer clientId=consumer-1, groupId=testTopic-002] Discovered group coordinator xxxxx.com:9092 (id: 2147483647 rack: null)
[28 10:20:08,931 INFO ] [main] internals.ConsumerCoordinator - [Consumer clientId=consumer-1, groupId=testTopic-002] Revoking previously assigned partitions []
[28 10:20:08,931 INFO ] [main] internals.AbstractCoordinator - [Consumer clientId=consumer-1, groupId=testTopic-002] (Re-)joining group
[28 10:20:08,958 INFO ] [main] internals.AbstractCoordinator - [Consumer clientId=consumer-1, groupId=testTopic-002] Successfully joined group with generation 1
[28 10:20:08,960 INFO ] [main] internals.ConsumerCoordinator - [Consumer clientId=consumer-1, groupId=testTopic-002] Setting newly assigned partitions [testTopic-0]
[28 10:20:08,978 INFO ] [main] internals.Fetcher - [Consumer clientId=consumer-1, groupId=testTopic-002] Resetting offset for partition testTopic-0 to offset 0.
topic = testTopic,partition = 0,offset = 0, key = H10, value = {"accessId":"123456","accessName":"源码婆媳10","busScope":"01","iconUrl":"http://www.baidu.com"}
topic = testTopic,partition = 0,offset = 1, key = H11, value = {"accessId":"123456","accessName":"源码婆媳11","busScope":"01","iconUrl":"http://www.baidu.com"}
topic = testTopic,partition = 0,offset = 2, key = H12, value = {"accessId":"123456","accessName":"源码婆媳12","busScope":"01","iconUrl":"http://www.baidu.com"}
topic = testTopic,partition = 0,offset = 3, key = H13, value = {"accessId":"123456","accessName":"源码婆媳13","busScope":"01","iconUrl":"http://www.baidu.com"}
topic = testTopic,partition = 0,offset = 4, key = H14, value = {"accessId":"123456","accessName":"源码婆媳14","busScope":"01","iconUrl":"http://www.baidu.com"}
topic = testTopic,partition = 0,offset = 5, key = H15, value = {"accessId":"123456","accessName":"源码婆媳15","busScope":"01","iconUrl":"http://www.baidu.com"}
topic = testTopic,partition = 0,offset = 6, key = H16, value = {"accessId":"123456","accessName":"源码婆媳16","busScope":"01","iconUrl":"http://www.baidu.com"}
topic = testTopic,partition = 0,offset = 7, key = H17, value = {"accessId":"123456","accessName":"源码婆媳17","busScope":"01","iconUrl":"http://www.baidu.com"}
topic = testTopic,partition = 0,offset = 8, key = H18, value = {"accessId":"123456","accessName":"源码婆媳18","busScope":"01","iconUrl":"http://www.baidu.com"}
topic = testTopic,partition = 0,offset = 9, key = H19, value = {"accessId":"123456","accessName":"源码婆媳19","busScope":"01","iconUrl":"http://www.baidu.com"}

使用低版本（0.8.22）客户端代码

public class SimpleConsumerExample {
 
    private static kafka.javaapi.consumer.ConsumerConnector consumer;
 
    public static void consume() {
 
        Properties props = new Properties();
        // zookeeper 配置
        props.put("zookeeper.connect", "127.0.0.1:2181");
 
        // group 代表一个消费组
        props.put("group.id", "jd-group");
 
        // zk连接超时
        props.put("zookeeper.session.timeout.ms", "4000");
        props.put("zookeeper.sync.time.ms", "200");
        props.put("auto.commit.interval.ms", "1000");
        props.put("auto.offset.reset", "smallest");
        // 序列化类
        props.put("serializer.class", "kafka.serializer.StringEncoder");
 
        ConsumerConfig config = new ConsumerConfig(props);
 
        consumer = kafka.consumer.Consumer.createJavaConsumerConnector(config);
 
        Map<String, Integer> topicCountMap = new HashMap<String, Integer>();
        topicCountMap.put("testTopic", new Integer(1));
 
        StringDecoder keyDecoder = new StringDecoder(new VerifiableProperties());
        StringDecoder valueDecoder = new StringDecoder(new VerifiableProperties());
 
        Map<String, List<KafkaStream<String, String>>> consumerMap = consumer.createMessageStreams(topicCountMap,
                keyDecoder, valueDecoder);
        KafkaStream<String, String> stream = consumerMap.get("testTopic").get(0);
        ConsumerIterator<String, String> it = stream.iterator();
        while (it.hasNext())
            System.out.println(it.next().message());
    }
 
    public static void main(String[] args) {
        consume();
    }
}

消费者运行后，一直消费不到消息

[28 09:51:41,590 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] producer.SyncProducer - Connected to xxxxxx.xx.com:9092 for producing
[28 09:51:41,591 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] producer.SyncProducer - Disconnecting from xxxxxx.xx.com:9092
[28 09:51:41,592 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] consumer.ConsumerFetcherManager - [ConsumerFetcherManager-1566957087042] Added fetcher for partitions ArrayBuffer([[testTopic,0], initOffset 0 to broker id:0,host:xxxxxx.xx.com,port:9092] )
[28 09:51:41,798 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] utils.VerifiableProperties - Verifying properties
[28 09:51:41,798 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] utils.VerifiableProperties - Property client.id is overridden to jd-group
[28 09:51:41,798 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] utils.VerifiableProperties - Property metadata.broker.list is overridden to xxxxxx.xx.com:9092
[28 09:51:41,798 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] utils.VerifiableProperties - Property request.timeout.ms is overridden to 30000
[28 09:51:41,798 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] client.ClientUtils$ - Fetching metadata from broker id:0,host:xxxxxx.xx.com,port:9092 with correlation id 65 for 1 topic(s) Set(testTopic)
[28 09:51:41,799 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] producer.SyncProducer - Connected to xxxxxx.xx.com:9092 for producing
[28 09:51:41,824 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] producer.SyncProducer - Disconnecting from xxxxxx.xx.com:9092
[28 09:51:41,825 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] consumer.ConsumerFetcherManager - [ConsumerFetcherManager-1566957087042] Added fetcher for partitions ArrayBuffer([[testTopic,0], initOffset 0 to broker id:0,host:xxxxxx.xx.com,port:9092] )
[28 09:51:42,032 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] utils.VerifiableProperties - Verifying properties
[28 09:51:42,032 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] utils.VerifiableProperties - Property client.id is overridden to jd-group
[28 09:51:42,032 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] utils.VerifiableProperties - Property metadata.broker.list is overridden to xxxxxx.xx.com:9092
[28 09:51:42,032 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] utils.VerifiableProperties - Property request.timeout.ms is overridden to 30000
[28 09:51:42,033 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] client.ClientUtils$ - Fetching metadata from broker id:0,host:xxxxxx.xx.com,port:9092 with correlation id 66 for 1 topic(s) Set(testTopic)
[28 09:51:42,035 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] producer.SyncProducer - Connected to xxxxxx.xx.com:9092 for producing
[28 09:51:42,041 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] producer.SyncProducer - Disconnecting from xxxxxx.xx.com:9092
[28 09:51:42,041 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] consumer.ConsumerFetcherManager - [ConsumerFetcherManager-1566957087042] Added fetcher for partitions ArrayBuffer([[testTopic,0], initOffset 0 to broker id:0,host:xxxxxx.xx.com,port:9092] )
[28 09:51:42,251 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] utils.VerifiableProperties - Verifying properties
[28 09:51:42,252 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] utils.VerifiableProperties - Property client.id is overridden to jd-group
[28 09:51:42,252 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] utils.VerifiableProperties - Property metadata.broker.list is overridden to xxxxxx.xx.com:9092
[28 09:51:42,252 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] utils.VerifiableProperties - Property request.timeout.ms is overridden to 30000
[28 09:51:42,253 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] client.ClientUtils$ - Fetching metadata from broker id:0,host:xxxxxx.xx.com,port:9092 with correlation id 67 for 1 topic(s) Set(testTopic)
[28 09:51:42,254 INFO ] [jd-group_xxxxxx-1566957086889-abf02be8-leader-finder-thread] producer.SyncProducer - Connected to xxxxxx.xx.com:9092 for producing

Kafka Broker的server.log一直在刷错误日志：

[2019-08-28 09:51:42,045] ERROR [KafkaApi-0] Error when handling request {replica_id=-1,max_wait_time=100,min_bytes=1,topics=[{topic=testTopic,partitions=[{partition=0,fetch_offset=0,max_bytes=1048576}]}]} (kafka.server.KafkaApis)
java.lang.IllegalArgumentException: Magic v0 does not support record headers
    at org.apache.kafka.common.record.MemoryRecordsBuilder.appendWithOffset(MemoryRecordsBuilder.java:385)
    at org.apache.kafka.common.record.MemoryRecordsBuilder.append(MemoryRecordsBuilder.java:568)
    at org.apache.kafka.common.record.AbstractRecords.convertRecordBatch(AbstractRecords.java:117)
    at org.apache.kafka.common.record.AbstractRecords.downConvert(AbstractRecords.java:98)
    at org.apache.kafka.common.record.FileRecords.downConvert(FileRecords.java:245)
    at kafka.server.KafkaApis$$anonfun$kafka$server$KafkaApis$$convertedPartitionData$1$1$$anonfun$apply$5.apply(KafkaApis.scala:523)
    at kafka.server.KafkaApis$$anonfun$kafka$server$KafkaApis$$convertedPartitionData$1$1$$anonfun$apply$5.apply(KafkaApis.scala:521)
    at scala.Option.map(Option.scala:146)
    at kafka.server.KafkaApis$$anonfun$kafka$server$KafkaApis$$convertedPartitionData$1$1.apply(KafkaApis.scala:521)
    at kafka.server.KafkaApis$$anonfun$kafka$server$KafkaApis$$convertedPartitionData$1$1.apply(KafkaApis.scala:511)
    at scala.Option.flatMap(Option.scala:171)
    at kafka.server.KafkaApis.kafka$server$KafkaApis$$convertedPartitionData$1(KafkaApis.scala:511)
    at kafka.server.KafkaApis$$anonfun$kafka$server$KafkaApis$$createResponse$2$1.apply(KafkaApis.scala:559)
    at kafka.server.KafkaApis$$anonfun$kafka$server$KafkaApis$$createResponse$2$1.apply(KafkaApis.scala:558)
    at scala.collection.Iterator$class.foreach(Iterator.scala:891)
    at scala.collection.AbstractIterator.foreach(Iterator.scala:1334)
    at scala.collection.IterableLike$class.foreach(IterableLike.scala:72)
    at scala.collection.AbstractIterable.foreach(Iterable.scala:54)
    at kafka.server.KafkaApis.kafka$server$KafkaApis$$createResponse$2(KafkaApis.scala:558)
    at kafka.server.KafkaApis$$anonfun$kafka$server$KafkaApis$$fetchResponseCallback$1$1.apply$mcVI$sp(KafkaApis.scala:579)
    at kafka.server.ClientQuotaManager.recordAndThrottleOnQuotaViolation(ClientQuotaManager.scala:196)
    at kafka.server.KafkaApis.sendResponseMaybeThrottle(KafkaApis.scala:2012)
    at kafka.server.KafkaApis.kafka$server$KafkaApis$$fetchResponseCallback$1(KafkaApis.scala:578)
    at kafka.server.KafkaApis$$anonfun$kafka$server$KafkaApis$$processResponseCallback$1$1.apply$mcVI$sp(KafkaApis.scala:598)
    at kafka.server.ClientQuotaManager.recordAndThrottleOnQuotaViolation(ClientQuotaManager.scala:196)
    at kafka.server.ClientQuotaManager.recordAndMaybeThrottle(ClientQuotaManager.scala:188)
    at kafka.server.KafkaApis.kafka$server$KafkaApis$$processResponseCallback$1(KafkaApis.scala:597)
    at kafka.server.KafkaApis$$anonfun$handleFetchRequest$1.apply(KafkaApis.scala:614)
    at kafka.server.KafkaApis$$anonfun$handleFetchRequest$1.apply(KafkaApis.scala:614)
    at kafka.server.ReplicaManager.fetchMessages(ReplicaManager.scala:639)
    at kafka.server.KafkaApis.handleFetchRequest(KafkaApis.scala:606)
    at kafka.server.KafkaApis.handle(KafkaApis.scala:98)
    at kafka.server.KafkaRequestHandler.run(KafkaRequestHandler.scala:66)
    at java.lang.Thread.run(Thread.java:748)

通过错误日志信息结合源码，我们发现，在Broker拉取到Kakfa消息后，调用fetchResponseCallback回调方法，创建返回信息时，会校验消费者Api版本，如果低于当前Broker版本与向下转换消息

      def fetchResponseCallback(bandwidthThrottleTimeMs: Int) {
        def createResponse(requestThrottleTimeMs: Int): RequestChannel.Response = {
          val convertedData = new util.LinkedHashMap[TopicPartition, FetchResponse.PartitionData]
          fetchedPartitionData.asScala.foreach { case (tp, partitionData) =>
            convertedData.put(tp, convertedPartitionData(tp, partitionData))
          }
          val response = new FetchResponse(convertedData, 0)
          val responseStruct = response.toStruct(versionId)

          trace(s"Sending fetch response to client $clientId of ${responseStruct.sizeOf} bytes.")
          response.responseData.asScala.foreach { case (topicPartition, data) =>
            // record the bytes out metrics only when the response is being sent
            brokerTopicStats.updateBytesOut(topicPartition.topic, fetchRequest.isFromFollower, data.records.sizeInBytes)
          }

          val responseSend = response.toSend(responseStruct, bandwidthThrottleTimeMs + requestThrottleTimeMs,
            request.connectionId, request.header)
          RequestChannel.Response(request, responseSend)
        }

        if (fetchRequest.isFromFollower)
          sendResponseExemptThrottle(createResponse(0))
        else
          sendResponseMaybeThrottle(request, request.header.clientId, requestThrottleMs =>
            requestChannel.sendResponse(createResponse(requestThrottleMs)))
      }



    def convertedPartitionData(tp: TopicPartition, data: FetchResponse.PartitionData) = {


      replicaManager.getMagic(tp).flatMap { magic =>
        val downConvertMagic = {
          if (magic > RecordBatch.MAGIC_VALUE_V0 && versionId <= 1 && !data.records.hasCompatibleMagic(RecordBatch.MAGIC_VALUE_V0))
            Some(RecordBatch.MAGIC_VALUE_V0)
          else if (magic > RecordBatch.MAGIC_VALUE_V1 && versionId <= 3 && !data.records.hasCompatibleMagic(RecordBatch.MAGIC_VALUE_V1))
            Some(RecordBatch.MAGIC_VALUE_V1)
          else
            None
        }

        downConvertMagic.map { magic =>
          trace(s"Down converting records from partition $tp to message format version $magic for fetch request from $clientId")
          //在这里会对消息进行向下转换
          val converted = data.records.downConvert(magic, fetchRequest.fetchData.get(tp).fetchOffset)
          new FetchResponse.PartitionData(data.error, data.highWatermark, FetchResponse.INVALID_LAST_STABLE_OFFSET,
            data.logStartOffset, data.abortedTransactions, converted)
        }

      }.getOrElse(data)
    }

对消息转换，最后会调用MemoryRecordsBuilder的appendWithOffset，在此方法中做一些调用，如果调用不通过就会抛出异常，Magic v0 does not support record headers就是在此方法中抛出的。因为Magic v0 和Magic v1版本的消息格式中，不支持header

private Long appendWithOffset(long offset, boolean isControlRecord, long timestamp, ByteBuffer key,
                                  ByteBuffer value, Header[] headers) {
        try {
            if (isControlRecord != isControlBatch)
                throw new IllegalArgumentException("Control records can only be appended to control batches");

            if (lastOffset != null && offset <= lastOffset)
                throw new IllegalArgumentException(String.format("Illegal offset %s following previous offset %s " +
                        "(Offsets must increase monotonically).", offset, lastOffset));

            if (timestamp < 0 && timestamp != RecordBatch.NO_TIMESTAMP)
                throw new IllegalArgumentException("Invalid negative timestamp " + timestamp);

            if (magic < RecordBatch.MAGIC_VALUE_V2 && headers != null && headers.length > 0)
                throw new IllegalArgumentException("Magic v" + magic + " does not support record headers");

            if (baseTimestamp == null)
                baseTimestamp = timestamp;

            if (magic > RecordBatch.MAGIC_VALUE_V1) {
                appendDefaultRecord(offset, timestamp, key, value, headers);
                return null;
            } else {
                return appendLegacyRecord(offset, timestamp, key, value);
            }
        } catch (IOException e) {
            throw new KafkaException("I/O exception when writing to the append stream, closing", e);
        }
    }

下面我们将Broker的Kafka版本换成1.0

运行低版本消费者：

[28 14:26:23,068 INFO ] [jd-group_xx-1566973572731-a5b3105a-leader-finder-thread] consumer.ConsumerFetcherManager - [ConsumerFetcherManager-1566973572960] Added fetcher for partitions ArrayBuffer([[testTopic,0], initOffset -1 to broker id:0,host:xx.xx.com,port:9092] )
{"accessId":"123456","accessName":"源码婆媳10","busScope":"01","iconUrl":"http://www.baidu.com"}
{"accessId":"123456","accessName":"源码婆媳11","busScope":"01","iconUrl":"http://www.baidu.com"}
{"accessId":"123456","accessName":"源码婆媳12","busScope":"01","iconUrl":"http://www.baidu.com"}
{"accessId":"123456","accessName":"源码婆媳13","busScope":"01","iconUrl":"http://www.baidu.com"}
{"accessId":"123456","accessName":"源码婆媳14","busScope":"01","iconUrl":"http://www.baidu.com"}
{"accessId":"123456","accessName":"源码婆媳15","busScope":"01","iconUrl":"http://www.baidu.com"}
{"accessId":"123456","accessName":"源码婆媳16","busScope":"01","iconUrl":"http://www.baidu.com"}
{"accessId":"123456","accessName":"源码婆媳17","busScope":"01","iconUrl":"http://www.baidu.com"}
{"accessId":"123456","accessName":"源码婆媳18","busScope":"01","iconUrl":"http://www.baidu.com"}
{"accessId":"123456","accessName":"源码婆媳19","busScope":"01","iconUrl":"http://www.baidu.com"}

很奇怪，怎么升级下Broker版本就可以正常消费消息了呢？不是说好的v0版本消息格式不支持header吗？

看了下1.0版本的源码，发现在做消息向下转换的时候调用的不是MemoryRecordsBuilder，而是RecordsUtil的convertRecordBatch，当发现v0或v1版本时，直接忽略header信息，这样消费者就能正常消息消息了

    private static MemoryRecordsBuilder convertRecordBatch(byte magic, ByteBuffer buffer, RecordBatchAndRecords recordBatchAndRecords) {
        RecordBatch batch = recordBatchAndRecords.batch;
        final TimestampType timestampType = batch.timestampType();
        long logAppendTime = timestampType == TimestampType.LOG_APPEND_TIME ? batch.maxTimestamp() : RecordBatch.NO_TIMESTAMP;

        MemoryRecordsBuilder builder = MemoryRecords.builder(buffer, magic, batch.compressionType(),
                timestampType, recordBatchAndRecords.baseOffset, logAppendTime);
        for (Record record : recordBatchAndRecords.records) {
            // Down-convert this record. Ignore headers when down-converting to V0 and V1 since they are not supported
            if (magic > RecordBatch.MAGIC_VALUE_V1)
                builder.append(record);
            else
                builder.appendWithOffset(record.offset(), record.timestamp(), record.key(), record.value());
        }

        builder.close();
        return builder;
    }

总结：

​ 通过之前的分析，解决上面的错误有三种方法

1）升级消费端Api版本，升级到新版本后，支持header

2）升级Broker版本，1.0以上版本，向下转换时，会忽略到header信息

3）最后一种方式，也很简单，那就是生产者不增加header信息。因为我们项目中LOG-ID暂时不是必须的，我们选择了此种方式，等消费者端版本全部升级之后，再添加header信息