Kafka Connect

lihongbao802024-03-27 18:08

confluent官网：https://debezium.io/documentation/reference/stable/connectors/sqlserver.html

Debezium官网：https://debezium.io/documentation/reference/stable/connectors/sqlserver.html

一、什么是 Kafka Connect？

Kafka Connect 是 Apache Kafka® 的一个免费开源组件，可作为集中式数据中心，用于在数据库、键值存储、搜索索引和文件系统之间进行简单的数据集成。

您可以使用 Kafka Connect 在 Apache Kafka 和其他数据系统之间流式传输数据，并快速创建用于将大型数据集移入和移出 Kafka® 的连接器。

二、Kafka Connect 下载（sql server下载举例）

confluent官网下载地址：
https://www.confluent.io/hub/debezium/debezium-connector-sqlserver


三、Kafka Connect 启动

1、修改配置文件（文件在kafka的安装目录有）

vim /opt/kafka/connect-config/connect-distributed.properties

powershell 复制代码 
bootstrap.servers=192.168.26.25:9092
 
group.id=connect-cluster

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
 
plugin.path=/data/qys/infra/kafka_2.12-3.6.1/plugins
topic.creation.default.partitions=1
topic.creation.default.replication.factor=1

主要修改内容：

  • 指定好bootstrap server地址 bootstrap.servers

  • 默认分区与分片

    topic.creation.default.partitions=1

    topic.creation.default.replication.factor=1

  • 插件地址

    plugin.path=

2、下载插件

参照 二、Kafka Connect 下载

3、启动kafka-connect进程（前提kafka已经启动）

powershell 复制代码 
/opt/kafka/bin/connect-distributed.sh -daemon /opt/kafka/config/connect-distributed.properties

postman查看

四、docker-compose启动kafka-connect示例

1、修改配置文件（文件在kafka的安装目录有）

vim ./kafka-connect/conf/connect-distributed.properties(参照上文三的1)

2、下载sql server

参照 二、Kafka Connect 下载

3、docker-compose.yaml

powershell 复制代码 
version: '2.4'
services:
  zookeeper:
    image: wurstmeister/zookeeper
    container_name: zk
    ports:
      - "2181:2181"
    restart: always
    volumes:
      - ./zookeeper_data:/opt/zookeeper-3.4.13/data

  kafka1:
    image: wurstmeister/kafka:2.12-2.3.0
    container_name: kafka1
    ports:
      - "32771:9092" 
    environment:
      TZ: Asia/Shanghai  # 设置为所需的时区
      KAFKA_ADVERTISED_HOST_NAME: 192.168.180.46
      KAFKA_ZOOKEEPER_CONNECT: zookeeper:2181
      KAFKA_AUTO_CREATE_TOPICS_ENABLE: 'true'
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock
      - ./kafka_data/kafka1_data:/kafka
    restart: always
    depends_on:
      - zookeeper

  kafka2:
    image: wurstmeister/kafka:2.12-2.3.0
    container_name: kafka2
    ports:
      - "32772:9092" 
    environment:
      TZ: Asia/Shanghai  # 设置为所需的时区
      KAFKA_ADVERTISED_HOST_NAME: 192.168.180.46
      KAFKA_ZOOKEEPER_CONNECT: zookeeper:2181
      KAFKA_AUTO_CREATE_TOPICS_ENABLE: 'true'
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock
      - ./kafka_data/kafka2_data:/kafka
    restart: always
    depends_on:
      - zookeeper

  kafka3:
    image: wurstmeister/kafka:2.12-2.3.0
    container_name: kafka3
    ports:
      - "32773:9092" 
    environment:
      TZ: Asia/Shanghai  # 设置为所需的时区
      KAFKA_ADVERTISED_HOST_NAME: 192.168.180.46
      KAFKA_ZOOKEEPER_CONNECT: zookeeper:2181
      KAFKA_AUTO_CREATE_TOPICS_ENABLE: 'true'
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock
      - ./kafka_data/kafka3_data:/kafka
    restart: always
    depends_on:
      - zookeeper

  kafka-connect:
    image: wurstmeister/kafka:2.12-2.3.0
    container_name: connect
    ports:
      - "38083:8083"
    entrypoint:
      - /opt/kafka/bin/connect-distributed.sh 
      - /opt/kafka/connect-config/connect-distributed.properties
    volumes:
      - /etc/localtime:/usr/share/zoneinfo/Asia/Shanghai
      - /var/run/docker.sock:/var/run/docker.sock
      - ./kafka-connect/conf:/opt/kafka/connect-config
      - ./kafka-connect/plugins:/opt/bitnami/kafka/plugin
    restart: always
    depends_on:
      - zookeeper

  kafka-client:
    image: wurstmeister/kafka:2.12-2.3.0
    entrypoint:
      - tail
      - -f
      - /etc/hosts
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock
      - ./kafka-connect/conf:/opt/kafka/connect-config
      - ./kafka-connect/plugins:/opt/kafka/plugins
    restart: always
    depends_on:
      - zookeeper

  kafdrop:
    image: obsidiandynamics/kafdrop
    container_name: kafdrop
    restart: "no"
    ports:
      - "9000:9000"
    environment:
      KAFKA_BROKERCONNECT: "kafka1:9092,kafka2:9092,kafka3:9092"
      TZ: Asia/Shanghai  # 设置为所需的时区
    depends_on:
      - zookeeper
      - kafka1
      - kafka2
      - kafka3

五、postman创建SQLserver的connec

1、sqlserver数据库开启cdc

powershell 复制代码 
-- 开启该库cdc
USE 库名
EXEC sys.sp_cdc_enable_db 
GO

-- 开启表CDC
use 库名;
EXEC sys.sp_cdc_enable_table 
@source_schema = N'schema名', 
@source_name = N'表名', 
@role_name = NULL, 
@supports_net_changes = 1 
GO

-- 关闭数据库CDC
USE 库名;
GO
EXEC sys.sp_cdc_disable_db

-- 关闭表的CDC功能
USE  库名;
GO
    EXEC sys.sp_cdc_disable_table
    @source_schema = N'schema名',
    @source_name   = N'表名',
    @capture_instance = N'schema名_表名'
GO


-- 查看数据库开启cdc检查
SELECT name ,is_cdc_enabled FROM sys.databases WHERE is_cdc_enabled = 1

-- 查看表cdc开启情况
use Libby;
SELECT name ,is_tracked_by_cdc FROM sys.tables WHERE is_tracked_by_cdc = 1

2、创建sqlserver的connect

  • postman创建connect
powershell 复制代码 
{
    "name": "cdc_mdata_test_3",
    "config": {
        "connector.class": "io.debezium.connector.sqlserver.SqlServerConnector",
        "database.hostname": "192.168.180.44",
        "database.port": "1433",
        "database.user": "sdp",
        "database.password": "shared@123",
        "database.dbname": "Libby",
        "table.whitelist": "MappingData.test",
        "database.server.name": "cdc_mdata_test_3",
        "database.history.kafka.bootstrap.servers": "192.168.180.46:32771",
        "database.history.kafka.topic": "cdc_mdata_test_3",
        "transforms": "Reroute",
        "transforms.Reroute.type": "io.debezium.transforms.ByLogicalTableRouter",
        "transforms.Reroute.topic.regex": "cdc_mdata_(.*)",
        "transforms.Reroute.topic.replacement": "cdc_md_combine",
        "errors.tolerance": "all"
    }
}
  • postman查看创建的connect

六、connect的restapi

1、参考网址：

https://docs.confluent.io/platform/current/connect/references/restapi.html

2、常用restapi举例

info method url
connect Cluster get http://192.168.180.46:38083/
connectors get http://192.168.180.46:38083/connectors
expand=status&expand=info get http://192.168.180.46:38083/connectors?expand=status\&expand=info
create-connector post http://192.168.180.46:38083/connectors
delete-connector delete http://192.168.180.46:38083/connectors/cdc_mdata_test_3(connector名称)
connector-config get http://192.168.180.46:38083/connectors/cdc_mdata_test_3(connector名称)/config
connector-status get http://192.168.180.46:38083/connectors/cdc_mdata_test_3(connector名称)/status
connector-pause put http://192.168.180.46:38083/connectors/cdc_mdata_test_3/pause
connector-resume put http://192.168.180.46:38083/connectors/cdc_mdata_test_3/resume
connector-stop put http://192.168.180.46:38083/connectors/mysql-connector-test46/stop
task-restart post http://192.168.180.46:38083/connectors/cdc_mdata_test_3/tasks/0/restart
task-status get http://192.168.180.46:38083/connectors/cdc_mdata_test_3/tasks/0/status
相关推荐
富士康质检员张全蛋4 天前
Kafka 事物
kafka
wear工程师4 天前
Kafka 消费者心跳正常，为什么还会被踢出组？拆清 max.poll.interval.ms
java·kafka
成为你的宁宁5 天前
【Kafka KRaft 集群】
分布式·kafka
2601_960906726 天前
地球或逃逸到相对安全的日心轨道
kafka·hbase·flume·storm
creator_Li6 天前
Kafka深入刨析-Consumer
分布式·kafka
JHC0000008 天前
基于 DrissionPage 与 Xvfb 的分布式无头爬虫架构
分布式·kafka
java1234_小锋8 天前
【免费】基于Spark实时医疗健康数据监测与疾病预测系统(Java版本+可视化大屏+Kafka+SpringBoot+Vue3) 锋哥原创出品，必属精品
大数据·spark·kafka·实时医疗健康数据监测·实时疾病预测系统
菜地里的小菜鸟9 天前
k8s部署单节点 Kafka（KRaft 模式）并启用 SASL 认证
kafka·kraft·k8s部署单节点kafkakraft模式·k8s部署kafka
Volunteer Technology9 天前
Kafka的消费全流程
分布式·kafka·linq
篮框坏了9 天前
Kafka 消费者卡死排查：一个 topic 堵了 18 小时，监控全绿没人发现
kafka
热门推荐
01GitHub 镜像站点02VSCode+CodeX扩展，无法启动03国内可直接用、免费额度/永久免费的大模型API清单（含 SiliconFlow、火山、阿里、智谱、百度、Kimi、DeepSeek、DMXAPI 等）042026 国产 AI 大模型横评：DeepSeek、通义千问、Kimi、文心一言、星火、豆包谁更能打？05微信历史版本含下载地址（ Windows PC | 安卓 | MAC ）及设置微信不更新06AI 编程 IDE 全景解析 2026：Agent 全面接管开发链路07Agnes AI 免费 API 接入指南：文本、生图、生视频，一套接口全免费08历年考研数学一、数学二、数学三真题试卷及答案PDF09如何新建文件夹？ 电脑新建文件夹的4种方法102026年7月AI圈大地震：GPT-5.6被政府限制、Claude入驻Slack、Anthropic自研芯片