如何通过Java在Kafka中创建主题

AdminUtils API已弃用。有一个新的API AdminZkClient，我们可以使用它来管理Kafka服务器中的主题。

String zookeeperHost = "127.0.0.1:2181";
Boolean isSucre = false;
int sessionTimeoutMs = 200000;
int connectionTimeoutMs = 15000;
int maxInFlightRequests = 10;
Time time = Time.SYSTEM;
String metricGroup = "myGroup";
String metricType = "myType";
KafkaZkClient zkClient = KafkaZkClient.apply(zookeeperHost,isSucre,sessionTimeoutMs,
                connectionTimeoutMs,maxInFlightRequests,time,metricGroup,metricType);

AdminZkClient adminZkClient = new AdminZkClient(zkClient);

String topicName1 = "myTopic";
int partitions = 3;
int replication = 1;
Properties topicConfig = new Properties();

adminZkClient.createTopic(topicName1,partitions,replication,
            topicConfig,RackAwareMode.Disabled$.MODULE$);

您可以参考此链接以获取详细信息：https : //www.analyticshut.com/streaming-services/kafka/create-and-list-kafka-topics-in-java/

在最新的（2.1.0）API中，该过程似乎已大大简化。使用适用于Kafka 2.1.0的最新API，可以按以下步骤进行操作

import org.apache.kafka.clients.admin.AdminClient;
import org.apache.kafka.clients.admin.CreateTopicsResult;
import org.apache.kafka.clients.admin.NewTopic;

Properties properties = new Properties();
properties.load(new FileReader(new File("kafka.properties")));

AdminClient adminClient = AdminClient.create(properties);
NewTopic newTopic = new NewTopic("topicName", 1, (short)1); //new NewTopic(topicName, numPartitions, replicationFactor)

List<NewTopic> newTopics = new ArrayList<NewTopic>();
newTopics.add(newTopic);

adminClient.createTopics(newTopics);
adminClient.close();

kafka.properties文件的内容如下

bootstrap.servers=localhost:9092
group.id=test
enable.auto.commit=true
auto.commit.interval.ms=1000
key.deserializer=org.apache.kafka.common.serialization.StringDeserializer
value.deserializer=org.apache.kafka.common.serialization.StringDeserializer

请注意，必须关闭AdminClient的实例才能反映新创建的主题。

我修好了..经过长时间的研究......

ZkClient zkClient = new ZkClient("localhost:2181", 10000, 10000);
AdminUtils.createTopic(zkClient, myTopic, 10, 1, new Properties());

从上面的代码中,ZkClient将创建一个主题,但是这个主题信息将无法识别kafka.所以我们要做的是,我们需要以下列方式为ZkClient创建对象,

首先导入以下声明,

import kafka.utils.ZKStringSerializer$;

并按以下方式为ZkClient创建对象,

ZkClient zkClient = new ZkClient("localhost:2181", 10000, 10000, ZKStringSerializer$.MODULE$);
AdminUtils.createTopic(zkClient, myTopic, 10, 1, new Properties());

编辑1 :(对于@ajkret评论)

由于api已被更改,上面的代码不适用于kafka> 0.9,使用以下代码为kafka> 0.9

import java.util.Properties;
import kafka.admin.AdminUtils;
import kafka.utils.ZKStringSerializer$;
import kafka.utils.ZkUtils;
import org.I0Itec.zkclient.ZkClient;
import org.I0Itec.zkclient.ZkConnection;

public class KafkaTopicCreationInJava
{
    public static void main(String[] args) throws Exception {
        ZkClient zkClient = null;
        ZkUtils zkUtils = null;
        try {
            String zookeeperHosts = "192.168.20.1:2181"; // If multiple zookeeper then -> String zookeeperHosts = "192.168.20.1:2181,192.168.20.2:2181";
            int sessionTimeOutInMs = 15 * 1000; // 15 secs
            int connectionTimeOutInMs = 10 * 1000; // 10 secs

            zkClient = new ZkClient(zookeeperHosts, sessionTimeOutInMs, connectionTimeOutInMs, ZKStringSerializer$.MODULE$);
            zkUtils = new ZkUtils(zkClient, new ZkConnection(zookeeperHosts), false);

            String topicName = "testTopic";
            int noOfPartitions = 2;
            int noOfReplication = 3;
            Properties topicConfiguration = new Properties();

            AdminUtils.createTopic(zkUtils, topicName, noOfPartitions, noOfReplication, topicConfiguration);

        } catch (Exception ex) {
            ex.printStackTrace();
        } finally {
            if (zkClient != null) {
                zkClient.close();
            }
        }
    }
}

对于那些试图在kafka v0.10.2.1中实现这一目的并遇到序列化错误的问题' java.io.StreamCorruptedException: invalid stream header: 3139322E'以下是带有必要导入的示例工作代码.

import org.I0Itec.zkclient.ZkClient;
import org.I0Itec.zkclient.ZkConnection;
import org.I0Itec.zkclient.exception.ZkMarshallingError;
import org.I0Itec.zkclient.serialize.ZkSerializer;
import org.apache.kafka.clients.consumer.KafkaConsumer;
import org.apache.kafka.common.PartitionInfo;

import kafka.admin.AdminUtils;
import kafka.admin.RackAwareMode;
import kafka.utils.ZKStringSerializer;
import kafka.utils.ZkUtils;

public static void createTopic(String topicName, int numPartitions, int numReplication) {
        ZkClient zkClient = null;
        ZkUtils zkUtils = null;
        try {
            String zookeeperHosts = "199.98.916.902:2181"; // If multiple zookeeper then -> String zookeeperHosts = "192.168.20.1:2181,192.168.20.2:2181";
            int sessionTimeOutInMs = 15 * 1000; // 15 secs
            int connectionTimeOutInMs = 10 * 1000; // 10 secs

            zkClient = new ZkClient(zookeeperHosts, sessionTimeOutInMs, connectionTimeOutInMs);
            //Ref: https://gist.github.com/jjkoshy/3842975
            zkClient.setZkSerializer(new ZkSerializer() {
                @Override
                public byte[] serialize(Object o) throws ZkMarshallingError {
                    return ZKStringSerializer.serialize(o);
                }

                @Override
                public Object deserialize(byte[] bytes) throws ZkMarshallingError {
                    return ZKStringSerializer.deserialize(bytes);
                }
            });

            zkUtils = new ZkUtils(zkClient, new ZkConnection(zookeeperHosts), false);

            int noOfPartitions = 2;
            int noOfReplication = 3;
            Properties topicConfiguration = new Properties();

            AdminUtils.createTopic(zkUtils, topicName, noOfPartitions, noOfReplication, topicConfiguration,
                    RackAwareMode.Enforced$.MODULE$);

        } catch (Exception ex) {
            ex.printStackTrace();
        } finally {
            if (zkClient != null) {
                zkClient.close();
            }
        }
    }

只是指向任何使用Kafka更新版本查看此内容的人(在撰写本文时,我使用的是Kafka v0.10.0.0).

你必须改变;

AdminUtils.createTopic(zkUtils, topicName, noOfPartitions, noOfReplications, topicConfiguration);

以下;

AdminUtils.createTopic(zkUtils, topicName, noOfPartitions, noOfReplications, true, Enforced$.MODULE$);

完成后关闭连接也是个好主意;

zkClient.close();