总体架构

postgresql

postgresql.conf

listen_addresses = '*'
shared_buffers = 128MB
dynamic_shared_memory_type = posix
wal_level = logical
max_wal_size = 1GB
min_wal_size = 80MB
max_wal_senders = 8
wal_keep_segments = 4
max_replication_slots = 4

pg_hba.conf

pg_hba.conf

重启PG

service postgresql restart

zk和kafka

下载
https://mirrors.aliyun.com/apache/
应该下载可执行文件的压缩包apache-zookeeper-x.x.x-bin.tar.gz，apache-zookeeper-x.x.x.tar.gz是源码
然后解压至指定目录

tar -xzvf

Debezium

下载对应数据库的Debezium连接器
https://debezium.io/documentation/reference/1.0/install.html
解压将jar拷贝到kafka的libs目录下

image.png

配置

服务器eth0_ip为172.18.4.109，公网ip为8.129.215.124

kafka

mkdir /usr/local/zgxsoft/etl/kafka/log

connect-distributed.properties

bootstrap.servers=172.18.4.109:9092
key.converter=org.apache.kafka.connect.json.JsonConverter
value.converter=org.apache.kafka.connect.json.JsonConverter
key.converter.schemas.enable=true
value.converter.schemas.enable=true
offset.storage.topic=connect-offsets
offset.storage.replication.factor=1
config.storage.topic=connect-configs
config.storage.replication.factor=1
status.storage.topic=connect-status
status.storage.replication.factor=1
offset.flush.interval.ms=10000

server.properties

broker.id=0
listeners=PLAINTEXT://172.18.4.109:9092
host.name = 172.18.4.109
advertised.listeners=PLAINTEXT://8.129.215.124:9092
num.network.threads=3
socket.send.buffer.bytes=102400
socket.receive.buffer.bytes=102400      
socket.request.max.bytes=104857600
log.dirs=/usr/local/zgxsoft/etl/kafka/log
num.partitions=1
num.recovery.threads.per.data.dir=1
such as 3.
offsets.topic.replication.factor=1
transaction.state.log.replication.factor=1
transaction.state.log.min.isr=1    
zookeeper.connect=172.18.4.109:2181

zk

mkdir /usr/local/zgxsoft/etl/zk/log
mkdir /usr/local/zgxsoft/etl/zk/data

cp zookeeper-3.4.14/conf/zoo_sample.cfg zookeeper-3.4.14/conf/zoo.cfg

zoo.cfg

tickTime=2000
initLimit=10
syncLimit=5
dataDir=/usr/local/zgxsoft/etl/zk/log
dataLogDir=/usr/local/zgxsoft/etl/zk/data
clientPort=2181

kafka connector

增加postgresql connector

postman发送post请求示例

启动

启动zk

sh zkServer.sh start

启动kafaka connector

sh connect-distributed.sh ../../config/connect-distributed.properties

启动kafaka

./kafka-server-start.sh -daemon config/server.properties

springboot消费端

public class MQDict {
    public static final String MQ_ADDRESS_COLLECTION="8.129.215.124:9092";          //kafka集群地址
    public static final String CONSUMER_TOPIC = "track_iqifu.public.track_iqifu";   //消费者连接的topic
    public static final String CONSUMER_GROUP_ID = "iqifu_group_001";               //groupId，可以分开配置
    public static final String CONSUMER_ENABLE_AUTO_COMMIT = "true";                //是否自动提交（消费者）
    /*
    消费者自动提交偏移量。enable.auto.commit 设为 true，那么每过一定时间间隔，消费者会自动把从 poll() 方法接收到的最大偏移量提交上去。
    提交时间间隔由 auto.commit.interval.ms 控制，默认是5s。
     */
    public static final String CONSUMER_AUTO_COMMIT_INTERVAL_MS = "1000";
    public static final String CONSUMER_SESSION_TIMEOUT_MS = "30000";               //连接超时时间
    public static final int CONSUMER_MAX_POLL_RECORDS = 10;                         //每次拉取数
    public static final Duration CONSUMER_POLL_TIME_OUT = Duration.ofMillis(6000);  //拉去数据超时时间
    /*
    指定了消费者在读取一个没有偏移量（offset）的分区或者偏移量无效的情况下（因消费者长时间失效，包含偏移量的记录已经过时井被删除）该作何处理，
    默认值是 latest，表示在 offset 无效的情况下，消费者将从最新的记录开始读取数据（在消费者启动之后生成的记录）。
    另一个值是 earliest，消费者将从起始位置读取分区的记录。
     */
    public static final String AUTO_OFFSET_RESET = "earliest";
    //client.id是用户特定的字符串，用来在每次请求中帮助跟踪调用。它应该可以逻辑上确认产生这个请求的应用
    public static final String CLIENT_ID = "client_1";

}

public class MQConsumer {
    private Logger logger = LoggerFactory.getLogger(getClass());

    private static KafkaConsumer consumer;

    /**
     *  初始化消费者
     */
    static {
        Properties configs = initConfig();
        consumer = new KafkaConsumer(configs);
        /*
         创建了消费者之后，需要订阅 Topic，subscribe() 方法接受一个主题列表作为参数。
         subscribe() 也可以接收一个正则表达式，匹配多个主题（如果有新的名称匹配的主题创建，会立即触发一次再均衡，消费者就可以读取新添加的主题）
         */
        consumer.subscribe(Arrays.asList(MQDict.CONSUMER_TOPIC));
    }

    /**
     *  初始化配置
     */
    private static Properties initConfig(){
        Properties props = new Properties();
        props.put("bootstrap.servers", MQDict.MQ_ADDRESS_COLLECTION);
        props.put("group.id", MQDict.CONSUMER_GROUP_ID);
        props.put("enable.auto.commit", MQDict.CONSUMER_ENABLE_AUTO_COMMIT);
        props.put("auto.commit.interval.ms", MQDict.CONSUMER_AUTO_COMMIT_INTERVAL_MS);
        props.put("session.timeout.ms", MQDict.CONSUMER_SESSION_TIMEOUT_MS);
        props.put("max.poll.records", MQDict.CONSUMER_MAX_POLL_RECORDS);
        props.put("auto.offset.reset", MQDict.AUTO_OFFSET_RESET);
        props.put("client.id", MQDict.CLIENT_ID);
        // key反序列化
        props.put("key.deserializer", StringDeserializer.class.getName());
        // value反序列化
        props.put("value.deserializer", StringDeserializer.class.getName());
        return props;
    }

    public void exeConsumer(IqifuBakDAO iqifuBakDAO) {
        try {
            while (true) {
                /*
                 poll() 返回一个记录列表。
                 每条记录都包含了记录所属主题的信息、记录所在分区的信息、记录在分区里的偏移量，以及记录的键值对。
                 */
                ConsumerRecords records = consumer.poll(MQDict.CONSUMER_POLL_TIME_OUT);
                if (records != null) {
                    records.forEach((ConsumerRecord record) -> {
                         //打印偏移量，key，value
                         System.out.printf("offset = %d, value = %s", record.offset(), record.key(),record.value());
                         //插入数据库
                         iqifuBakDAO.insert(record);
                    });
                }
            }
        }finally {
            // 关闭消费者,网络连接和 socket 也会随之关闭，并立即触发一次再均衡
            consumer.close();
        }
    }
}

/**
 * Spring boot监听类，用于启动Spring boot容器后加载数据
 */
@Component
public class ApplicationStartup implements ApplicationListener {
    private Logger logger = LoggerFactory.getLogger(getClass());

    @SuppressWarnings("all")
    @Autowired
    IqifuBakDAO iqifuBakDAO;

    @Override
    public void onApplicationEvent(ContextRefreshedEvent contextRefreshedEvent) {
        logger.info("#########################################注入DAO层接口，开始消费#########################################");
        new MQConsumer().exeConsumer(iqifuBakDAO);
    }
}

参考资料：

https://debezium.io/documentation/reference/1.1/connectors/postgresql.html
https://blog.csdn.net/liu_c_y/article/details/103537490?utm_medium=distribute.pc_relevant.none-task-blog-BlogCommendFromMachineLearnPai2-3.nonecase&depth_1-utm_source=distribute.pc_relevant.none-task-blog-BlogCommendFromMachineLearnPai2-3.nonecase
https://blog.csdn.net/u012551524/article/details/82798066/

Debezium:数据实时采集从Postgresql到Kafka

postgresql

zk和kafka

Debezium

配置

kafka

zk

kafka connector

启动

启动zk

启动kafaka connector

启动kafaka

springboot消费端

参考资料：

你可能感兴趣的:(Debezium:数据实时采集从Postgresql到Kafka)