使用Kafka Java API的Avro序列化器和反序列化器

Question

Kafka Avro序列化器和解串器不起作用。 我尝试使用kafka控制台使用者使用消息，并且可以看到消息已发布。

public class AvroProducer<T> {

    private static Properties props;
    static {
        props = new Properties();
        props.put(ProducerConfig.BOOTSTRAP_SERVERS_CONFIG,"localhost:9092");
        props.put(ProducerConfig.KEY_SERIALIZER_CLASS_CONFIG, StringSerializer.class);
        props.put(ProducerConfig.VALUE_SERIALIZER_CLASS_CONFIG, ByteArraySerializer.class);
        props.put(ProducerConfig.CLIENT_ID_CONFIG, "Kafka Avro Producer");
    }

    private static KafkaProducer producer = new KafkaProducer<>(props);

    public byte[] createRecords(T pojo) throws IOException{
        Schema.Parser parser = new Schema.Parser();
        Schema schema = null;
        try {
            schema = parser.parse(AvroProducer.class.getClassLoader().getResourceAsStream("syslog.avsc"));
        } catch (IOException e) {
            System.out.println(e.getLocalizedMessage());
        }

        final GenericData.Record record = new GenericData.Record(schema);
        schema.getFields().forEach(r -> record.put(r.name(), PropertyAccessorFactory.forDirectFieldAccess(pojo).getPropertyValue(r.name())));
        SpecificDatumWriter<GenericRecord> writer = new SpecificDatumWriter<>(schema);
        try(ByteArrayOutputStream os = new ByteArrayOutputStream()){
            BinaryEncoder encoder = EncoderFactory.get().binaryEncoder(os, null);

            writer.write(record, encoder);
            encoder.flush();
            byte[] bytemessage = os.toByteArray();
            return bytemessage;
        }

    }

    public static void sendMessage(byte[] bytemessage){
        ProducerRecord precord = new ProducerRecord<StringSerializer, byte[]>("jason", bytemessage);
        producer.send(precord);
    }
}



    public class AvroConsumer {

    private static Properties kafkaProps;

    static {
        kafkaProps = new Properties();
        kafkaProps.put(ConsumerConfig.AUTO_OFFSET_RESET_CONFIG, "earliest");
        kafkaProps.put("bootstrap.servers", "localhost:9092");
        kafkaProps.put(ConsumerConfig.KEY_DESERIALIZER_CLASS_CONFIG, StringDeserializer.class);
        kafkaProps.put(ConsumerConfig.VALUE_DESERIALIZER_CLASS_CONFIG, ByteArrayDeserializer.class);
        kafkaProps.put(ConsumerConfig.GROUP_ID_CONFIG, "AvroConsumer-GroupOne");
    }

    @SuppressWarnings("unchecked")
    public static void recieveRecord() throws IOException{
        try (KafkaConsumer<String,byte[]> kafkaConsumer = new KafkaConsumer<>(kafkaProps)) {
            kafkaConsumer.subscribe(Arrays.asList("jason"));
            while (true) {
                ConsumerRecords<String,byte[]> records = kafkaConsumer.poll(100);
                Schema.Parser parser = new Schema.Parser();
                final Schema schema = parser.parse(AvroProducer.class.getClassLoader().getResourceAsStream("syslog.avsc"));
                records.forEach(record -> {
                    SpecificDatumReader<GenericRecord> datumReader = new SpecificDatumReader<>(schema);
                    ByteArrayInputStream is = new ByteArrayInputStream(record.value());
                    BinaryDecoder binaryDecoder = DecoderFactory.get().binaryDecoder(is, null);
                    try {
                        Syslog log = (Syslog) datumReader.read(null, binaryDecoder);

                        System.out.println("Value: " + log);
                    } catch (IOException e) {
                        e.printStackTrace();
                    }
                });
            }
        }
    }

}

堆栈跟踪如下所示，请查找详细信息。 有人可以指导我正确的实施方式。 问题似乎出在铸造记录上。如何获取价值。 如何使用“特定基准读取器”读取数据。

Exception in thread "Thread-1" java.lang.ClassCastException: org.apache.avro.generic.GenericData$Record cannot be cast to com.cisco.daas.kafka.Syslog
at com.cisco.daas.kafka.AvroConsumer.lambda$0(AvroConsumer.java:46)
at java.lang.Iterable.forEach(Unknown Source)
at com.cisco.daas.kafka.AvroConsumer.recieveRecord(AvroConsumer.java:41)
at com.cisco.daas.kafka.MainApp$1.run(MainApp.java:32)
at java.lang.Thread.run(Unknown Source)

这是我要解析的架构

{
    "namespace": "com.example.syslogmessage",
    "type": "record",
    "name": "SysLogMessage",
    "fields": [{
            "name": "partyID",
            "type": "long"
        },
        {
            "name": "applianceID",
            "type": "string"
        },
        {
            "name": "message",
            "type": "string"
        },
        {
            "name": "severity",
            "type": "long"
        },
        {
            "name": "messageType",
            "type": "string"
        },
        {
            "name": "eventtime",
            "type": "long",
            "logicalType": "timestamp-millis"
        },
        {
            "name": "ipaddress",
            "type": "string"
        },
        {
            "name": "hostname",
            "type": "string"
        }
    ]
}

Answer 1

通过使用GenericRecord的键值可以解决此问题

public class AvroConsumer<T> {

    private static Properties kafkaProps;

    static {
        kafkaProps = new Properties();
        kafkaProps.put(ConsumerConfig.AUTO_OFFSET_RESET_CONFIG, "earliest");
        kafkaProps.put("bootstrap.servers", "localhost:9092");
        kafkaProps.put(ConsumerConfig.KEY_DESERIALIZER_CLASS_CONFIG, StringDeserializer.class);
        kafkaProps.put(ConsumerConfig.VALUE_DESERIALIZER_CLASS_CONFIG, ByteArrayDeserializer.class);
        kafkaProps.put(ConsumerConfig.GROUP_ID_CONFIG, "AvroConsumer-GroupOne");
    }

    public void recieveRecord() throws IOException {
        try (KafkaConsumer<String, byte[]> kafkaConsumer = new KafkaConsumer<>(kafkaProps)) {
            kafkaConsumer.subscribe(Arrays.asList("jason"));
            while (true) {
                ConsumerRecords<String, byte[]> records = kafkaConsumer.poll(100);
                Schema.Parser parser = new Schema.Parser();
                final Schema schema = parser
                        .parse(AvroProducer.class.getClassLoader().getResourceAsStream("syslog.avsc"));
                records.forEach(record -> {
                    SpecificDatumReader<T> datumReader = new SpecificDatumReader<>(schema);
                    ByteArrayInputStream is = new ByteArrayInputStream(record.value());
                    BinaryDecoder binaryDecoder = DecoderFactory.get().binaryDecoder(is, null);
                    try {
                        T log = datumReader.read(null, binaryDecoder);
                        System.out.println("Value: " + log);
                    } catch (IOException e) {
                        e.printStackTrace();
                    }
                });
            }
        }
    }

}