使用 KStream 将字符串更改为 avro inKafka Streams 时出现空指针异常

Question

我被这个问题困住了，我不知道发生了什么。我正在尝试使用 Kafka 流将日志写入主题。在另一端，我让 Kafka-connect 将每个条目输入 MySQL。所以，基本上我需要的是一个 Kafka 流程序，它将主题作为字符串读取并将其解析为 Avro 格式，然后将其输入到不同的主题中。

这是我写的代码：

        //Define schema
        String userSchema = "{"
                + "\"type\":\"record\","
                + "\"name\":\"myrecord\","
                + "\"fields\":["
                + "  { \"name\":\"ID\", \"type\":\"int\" },"
                + "  { \"name\":\"COL_NAME_1\", \"type\":\"string\" },"
                + "  { \"name\":\"COL_NAME_2\", \"type\":\"string\" }"
    + "]}";

        String key = "key1";
        Schema.Parser parser = new Schema.Parser();
        Schema schema = parser.parse(userSchema);

//Settings
       System.out.println("Kafka Streams Demonstration");
        //Settings
        Properties settings = new Properties();
        // Set a few key parameters
        settings.put(StreamsConfig.APPLICATION_ID_CONFIG, APP_ID);
        // Kafka bootstrap server (broker to talk to); ubuntu is the host name for my VM running Kafka, port 9092 is where the (single) broker listens
        settings.put(StreamsConfig.BOOTSTRAP_SERVERS_CONFIG, "localhost:9092");
        // Apache ZooKeeper instance keeping watch over the Kafka cluster; ubuntu is the host name for my VM running Kafka, port 2181 is where the ZooKeeper listens
        settings.put(StreamsConfig.ZOOKEEPER_CONNECT_CONFIG, "localhost:2181");
        // default serdes for serialzing and deserializing key and value from and to streams in case no specific Serde is specified
        settings.put(StreamsConfig.KEY_SERDE_CLASS_CONFIG, Serdes.String().getClass().getName());
        settings.put(StreamsConfig.VALUE_SERDE_CLASS_CONFIG, Serdes.String().getClass().getName());
        settings.put(StreamsConfig.STATE_DIR_CONFIG ,"/tmp");
        // to work around exception Exception in thread "StreamThread-1" java.lang.IllegalArgumentException: Invalid timestamp -1
        // at org.apache.kafka.clients.producer.ProducerRecord.<init>(ProducerRecord.java:60)
        // see: https://groups.google.com/forum/#!topic/confluent-platform/5oT0GRztPBo

        // Create an instance of StreamsConfig from the Properties instance
        StreamsConfig config = new StreamsConfig(getProperties());
        final Serde < String > stringSerde = Serdes.String();
        final Serde < Long > longSerde = Serdes.Long();
        final Serde<byte[]> byteArraySerde = Serdes.ByteArray();

        // building Kafka Streams Model                                                                                                                                                       
        KStreamBuilder kStreamBuilder = new KStreamBuilder();
        // the source of the streaming analysis is the topic with country messages                                                                                                            
        KStream<byte[], String> instream =
            kStreamBuilder.stream(byteArraySerde, stringSerde, "sqlin");

       final KStream<byte[], GenericRecord> outstream = instream.mapValues(new ValueMapper<String, GenericRecord>() {
            @Override
            public GenericRecord apply(final String record) {
                System.out.println(record);
                GenericRecord avroRecord = new GenericData.Record(schema);
                String[] array = record.split(" ", -1);
                for (int i = 0; i < array.length; i = i + 1) {
                    if (i == 0)
                        avroRecord.put("ID", Integer.parseInt(array[0]));
                    if (i == 1)
                        avroRecord.put("COL_NAME_1", array[1]);
                    if (i == 2)
                        avroRecord.put("COL_NAME_2", array[2]);
                }
                System.out.println(avroRecord);
                return avroRecord;
            }
          });
        outstream.to("sqlout");

这是我得到空指针异常后的输出：

java -cp streams-examples-3.2.1-standalone.jar io.confluent.examples.streams.sql
Kafka Streams Demonstration
Start
Now started CountriesStreams Example
5 this is
{"ID": 5, "COL_NAME_1": "this", "COL_NAME_2": "is"}
Exception in thread "StreamThread-1" java.lang.NullPointerException
    at org.apache.kafka.streams.processor.internals.SinkNode.process(SinkNode.java:81)
    at org.apache.kafka.streams.processor.internals.ProcessorContextImpl.forward(ProcessorContextImpl.java:83)
    at org.apache.kafka.streams.kstream.internals.KStreamMapValues$KStreamMapProcessor.process(KStreamMapValues.java:42)
    at org.apache.kafka.streams.processor.internals.ProcessorNode$1.run(ProcessorNode.java:48)
    at org.apache.kafka.streams.processor.internals.StreamsMetricsImpl.measureLatencyNs(StreamsMetricsImpl.java:188)
    at org.apache.kafka.streams.processor.internals.ProcessorNode.process(ProcessorNode.java:134)
    at org.apache.kafka.streams.processor.internals.ProcessorContextImpl.forward(ProcessorContextImpl.java:83)
    at org.apache.kafka.streams.processor.internals.SourceNode.process(SourceNode.java:70)
    at org.apache.kafka.streams.processor.internals.StreamTask.process(StreamTask.java:197)
    at org.apache.kafka.streams.processor.internals.StreamThread.runLoop(StreamThread.java:627)
    at org.apache.kafka.streams.processor.internals.StreamThread.run(StreamThread.java:361)

主题 sqlin 包含一些消息，由一个数字后跟两个单词组成。请注意两行打印：该函数获取一条消息，并在捕获空指针之前成功解析它。问题是我是 Java、Kafka 和 Avro 的新手，所以我不确定我哪里出错了。我是否正确设置了 Avro 架构？还是使用 kstream 错误？在这里的任何帮助将不胜感激。

Answer 1

我认为问题出在以下几行：

outstream.to("sqlout");

默认情况下，您的应用程序配置为使用String serde 作为记录键和记录值：

settings.put(StreamsConfig.KEY_SERDE_CLASS_CONFIG, Serdes.String().getClass().getName());
settings.put(StreamsConfig.VALUE_SERDE_CLASS_CONFIG, Serdes.String().getClass().getName());

由于outstream 的类型为KStream<byte[], GenericRecord>，因此在调用to() 时必须提供明确的serdes：

// sth like
outstream.to(Serdes.ByteArray(), yourGenericAvroSerde, "sqlout");

仅供参考：Confluent 平台的下一个版本（预计到达时间：本月 = 六月 2017) 将附带ready-to-use generic + specific Avro serde 与Confluent schema registry 集成。这 应该让你的生活更轻松。

请在https://stackoverflow.com/a/44433098/1743580 上查看我的回答以了解更多详情。