Kafka 产生多值消息

Question

I am new to working with Kafka .

Am I able to send a message with multiple values? ie, I have this example here which is producing a message with a random number, however I would like to add a timestamp for each produced record. What is the standard way to do this?

    while(true){
        for(int key=0; key < 10000; key++){
            Random rand = new Random();
            int  n = rand.nextInt(50) + 1;

            SimpleDateFormat formatter = new SimpleDateFormat("dd/MM/yyyy HH:mm:ss");
            Date date = new Date();
            ProducerRecord<String, String> producerRecord = new ProducerRecord<String, String>("java-topic", Integer.toString(1),Integer.toString(n));
            producer.send(producerRecord);
            Thread.sleep(10000);
        }
        //producer.close();
    }

Answer 1

Internally Kafka broker (server) only stores a byte array. Thus You will have to encode the your message to contain all values.
A popular method to do this is to use JSON encoding in message

int numbers[] = {1,2,3};
String msg = String.format("{\"numbers\": %s, \"timestamp\": \"%s\"}", 
                java.util.Arrays.toString(numbers), timestamp);
ProducerRecord<String, String> producerRecord = new ProducerRecord<String, String>(
       "java-topic", msg);

Note a timestamp is added by the kafka server when it receives a message which can be seen in ConsumerRecord.timestamp()

Answer 2

Thanks for the help!! Output answer

while(true){
    for(int key=0; key < 10000; key++){
        Random rand = new Random();
        Date date = new Date();
        SimpleDateFormat formatter = new SimpleDateFormat("dd/MM/yyyy HH:mm:ss");

        int  n = rand.nextInt(50) + 1;
        String today = formatter.format(date);
        String msg = String.format("{\"numbers\": %s, \"timestamp\": \"%s\"}", n, today);

        ProducerRecord<String, String> producerRecord = new ProducerRecord<String, String>(
                "java-topic", Integer.toString(1), msg);
        producer.send(producerRecord);
        Thread.sleep(1000);
    }
    //producer.close();
}