一文讲清楚FusionInsight MRS CDL如何使用

àì夳堔傛蜴生んèń 2022-09-14 01:50 328阅读 0赞

> **摘要：**CDL是一种简单、高效的数据实时集成服务，能够从各种OLTP数据库中抓取Data Change事件，然后推送至Kafka中，最后由Sink Connector消费Topic中的数据并导入到大数据生态软件应用中，从而实现数据的实时入湖。

本文分享自华为云社区《[华为FusionInsight MRS CDL使用指南][FusionInsight MRS CDL]》，作者：晋红轻。

## **说明** ##

CDL是一种简单、高效的数据实时集成服务，能够从各种OLTP数据库中抓取Data Change事件，然后推送至Kafka中，最后由Sink Connector消费Topic中的数据并导入到大数据生态软件应用中，从而实现数据的实时入湖。

CDL服务包含了两个重要的角色：CDLConnector和CDLService。CDLConnector是具体执行数据抓取任务的实例，CDLService是负责管理和创建任务的实例。

本此实践介绍以mysql作为数据源进行数据抓取

## 前提条件 ##

*  MRS集群已安装CDL服务。
 *  MySQL数据库需要开启mysql的bin log功能（默认情况下是开启的）。

查看MySQL是否开启bin log：

使用工具或者命令行连接MySQL数据库（本示例使用navicat工具连接），执行show variables like 'log\_%'命令查看。

例如在navicat工具选择"File > New Query"新建查询，输入如下SQL命令，单击"Run"在结果中"log\_bin"显示为"ON"则表示开启成功。

show variables like 'log_%'

![c9e2c85bc2b7bc70537f0fa9874383e4.png][]

## 工具准备 ##

现在cdl只能使用rest api的方式进行命令提交，所以需要提前安装工具进行调试。本文使用VSCode工具。

![7c09f9e4aa35d80a01b751b22a236e03.png][]

完成之后安装rest client插件：

![70205f4e05894dc1e7d1c3ccab25c3c1.png][]

完成之后创建一个cdl.http的文件进行编辑：

![4ea3e3ce00c406710790b443e29f832e.png][]

## 创建CDL任务 ##

CDL任务创建的流程图如下所示：

![21309232afc914b3cdaff88ff63ae6ff.png][]

说明：需要先创建一个MySQL link, 在创建一个Kafka link, 然后再创建一个CDL同步任务并启动。

MySQL link部分rest请求代码

@hostname = 172.16.9.113
    @port = 21495
    @host = {
       {hostname}}:{
       {port}}
    @bootstrap = "172.16.9.113:21007"
    @bootstrap_normal = "172.16.9.113:21005"
    @mysql_host = "172.16.2.118"
    @mysql_port = "3306"
    @mysql_database = "hudi"
    @mysql_user = "root"
    @mysql_password = "Huawei@123"
    
    ### get links
    get https://{
       {host}}/api/v1/cdl/link
    
    ### mysql link validate
    
    post https://{
       {host}}/api/v1/cdl/link?validate=true
    content-type: application/json
    
    {
    "name": "MySQL_link", //link名，全局唯一，不能重复
    "description":"MySQL connection", //link描述
    "link-type":"mysql", //link的类型
    "enabled":"true",
    "link-config-values":  {
    "inputs": [
            { "name": "host", "value": {
       {mysql_host}} }, //数据库安装节点的ip
            { "name": "port", "value": {
       {mysql_port}} },//数据库监听的端口
            { "name": "database.name", "value": {
       {mysql_database}} }, //连接的数据库名
            { "name": "user", "value": {
       {mysql_user}} }, //用户
            { "name": "password","value": {
       {mysql_password}} } ,//密码
            { "name":"schema", "value": {
       {mysql_database}}}//同数据库名
            ]
        }
    }
    
    ### mysql link create
    
    post https://{
       {host}}/api/v1/cdl/link
    content-type: application/json
    
    {
    "name": "MySQL_link", //link名，全局唯一，不能重复
    "description":"MySQL connection", //link描述
    "link-type":"mysql", //link的类型
    "enabled":"true",
    "link-config-values":  {
    "inputs": [
            { "name": "host", "value": {
       {mysql_host}} }, //数据库安装节点的ip
            { "name": "port", "value": {
       {mysql_port}} },//数据库监听的端口
            { "name": "database.name", "value": {
       {mysql_database}} }, //连接的数据库名
            { "name": "user", "value": {
       {mysql_user}} }, //用户
            { "name": "password","value": {
       {mysql_password}} } ,//密码
            { "name":"schema", "value": {
       {mysql_database}}}//同数据库名
            ]
        }
    }
    
    ### mysql link update
    
    put https://{
       {host}}/api/v1/cdl/link/MySQL_link
    content-type: application/json
    
    {
    "name": "MySQL_link", //link名，全局唯一，不能重复
    "description":"MySQL connection", //link描述
    "link-type":"mysql", //link的类型
    "enabled":"true",
    "link-config-values":  {
    "inputs": [
            { "name": "host", "value": {
       {mysql_host}} }, //数据库安装节点的ip
            { "name": "port", "value": {
       {mysql_port}} },//数据库监听的端口
            { "name": "database.name", "value": {
       {mysql_database}} }, //连接的数据库名
            { "name": "user", "value": {
       {mysql_user}} }, //用户
            { "name": "password","value": {
       {mysql_password}} } ,//密码
            { "name":"schema", "value": {
       {mysql_database}}}//同数据库名
            ]
        }
    }

Kafka link部分rest请求代码

### get links
    get https://{
       {host}}/api/v1/cdl/link
    
    ### kafka link validate
    
    post https://{
       {host}}/api/v1/cdl/link?validate=true
    content-type: application/json
    
    {
    "name": "kafka_link",
    "description":"test kafka link",
    "link-type":"kafka",
    "enabled":"true",
    "link-config-values":  {
    "inputs": [
            { "name": "bootstrap.servers", "value": "172.16.9.113:21007" },
            { "name": "sasl.kerberos.service.name", "value": "kafka" },
            { "name": "security.protocol","value": "SASL_PLAINTEXT" }//安全模式为SASL_PLAINTEXT，普通模式为PLAINTEXT
            ]
        }
    }
    
    ### kafka link create
    
    post https://{
       {host}}/api/v1/cdl/link
    content-type: application/json
    
    {
    "name": "kafka_link",
    "description":"test kafka link",
    "link-type":"kafka",
    "enabled":"true",
    "link-config-values":  {
    "inputs": [
            { "name": "bootstrap.servers", "value": "172.16.9.113:21007" },
            { "name": "sasl.kerberos.service.name", "value": "kafka" },
            { "name": "security.protocol","value": "SASL_PLAINTEXT" }//安全模式为SASL_PLAINTEXT，普通模式为PLAINTEXT
            ]
        }
    }
    
    ### kafka link update
    
    put https://{
       {host}}/api/v1/cdl/link/kafka_link
    content-type: application/json
    
    {
    "name": "kafka_link",
    "description":"test kafka link",
    "link-type":"kafka",
    "enabled":"true",
    "link-config-values":  {
    "inputs": [
            { "name": "bootstrap.servers", "value": "172.16.9.113:21007" },
            { "name": "sasl.kerberos.service.name", "value": "kafka" },
            { "name": "security.protocol","value": "SASL_PLAINTEXT" }//安全模式为SASL_PLAINTEXT，普通模式为PLAINTEXT
            ]
        }
    }

CDL任务命令部分rest请求代码

@hostname = 172.16.9.113
    @port = 21495
    @host = {
       {hostname}}:{
       {port}}
    @bootstrap = "172.16.9.113:21007"
    @bootstrap_normal = "172.16.9.113:21005"
    @mysql_host = "172.16.2.118"
    @mysql_port = "3306"
    @mysql_database = "hudi"
    @mysql_user = "root"
    @mysql_password = "Huawei@123"
    
    ### create job
    post https://{
       {host}}/api/v1/cdl/job
    content-type: application/json
    
    {
        "job_type": "CDL_JOB", //job类型,目前只支持CDL_JOB这一种
        "name": "mysql_to_kafka", //job名称
        "description":"mysql_to_kafka", //job描述
        "from-link-name": "MySQL_link",  //数据源Link
        "to-link-name": "kafka_link", //目标源Link
        "from-config-values": {
            "inputs": [
                {"name" : "connector.class", "value" : "com.huawei.cdc.connect.mysql.MysqlSourceConnector"},
                {"name" : "schema", "value" : "hudi"},
                {"name" : "db.name.alias", "value" : "hudi"},
                {"name" : "whitelist", "value" : "hudisource"},
                {"name" : "tables", "value" : "hudisource"},
                {"name" : "tasks.max", "value" : "10"},
                {"name" : "mode", "value" : "insert,update,delete"},
                {"name" : "parse.dml.data", "value" : "true"},
                {"name" : "schema.auto.creation", "value" : "false"},
                {"name" : "errors.tolerance", "value" : "all"},
                {"name" : "multiple.topic.partitions.enable", "value" : "false"},
                {"name" : "topic.table.mapping", "value" : "[
                        {\"topicName\":\"huditableout\", \"tableName\":\"hudisource\"}
                    ]"
                },
      	        {"name" : "producer.override.security.protocol", "value" : "SASL_PLAINTEXT"},//安全模式为SASL_PLAINTEXT，普通模式为PLAINTEXT
    	        {"name" : "consumer.override.security.protocol", "value" : "SASL_PLAINTEXT"}//安全模式为SASL_PLAINTEXT，普通模式为PLAINTEXT
            ]
        },
        "to-config-values": {"inputs": []},
        "job-config-values": {
            "inputs": [
                {"name" : "global.topic", "value" : "demo"}
            ]
        }
    }
    
    ### get all job
    get https://{
       {host}}/api/v1/cdl/job
    ### submit job
    put https://{
       {host}}/api/v1/cdl/job/mysql_to_kafka/start
    ### get job status
    get https://{
       {host}}/api/v1/cdl/submissions?jobName=mysql_to_kafka
    ### stop job
    put https://{
       {host}}/api/v1/cdl/job/mysql_to_kafka/submissions/13/stop
    ### delete job
    DELETE https://{
       {host}}/api/v1/cdl/job/mysql_to_kafka

## 场景验证 ##

生产库MySQL原始数据如下：

![91d0d74cba727f56a116253ce7d0567f.png][]

提交CDL任务之后

![492da3404aef7faaf310e160b5e2b21e.png][]

增加操作: insert into hudi.hudisource values (11,“蒋语堂”,38,“女”,“图”,“播放器”,28732);

对应kafka消息体：

![8c62d9a6c779436133517f9785fec7af.png][]

更改操作: UPDATE hudi.hudisource SET uname=‘Anne Marie333’ WHERE uid=11;

对应kafka消息体：

![addbdc9b0d8f0ef74751a47cb128be7c.png][]

删除操作：delete from hudi.hudisource where uid=11;

对应kafka消息体：

![4b2fed6793ffbf7343e977d7150c0a89.png][]

**[点击关注，第一时间了解华为云新鲜技术~][Link 1]**

[FusionInsight MRS CDL]: https://link.zhihu.com/?target=https%3A//bbs.huaweicloud.com/blogs/302537%3Futm_source%3Dzhihu%26utm_medium%3Dbbs-ex%26utm_campaign%3Dei%26utm_content%3Dcontent
[c9e2c85bc2b7bc70537f0fa9874383e4.png]: /images/20220828/0262073897624de2bc005c52bff27e4b.png
[7c09f9e4aa35d80a01b751b22a236e03.png]: /images/20220828/368c81ce927f44b0b3dd6e537b522bd6.png
[70205f4e05894dc1e7d1c3ccab25c3c1.png]: /images/20220828/5511022fe3e5467fbc92aef6ae64d17f.png
[4ea3e3ce00c406710790b443e29f832e.png]: /images/20220828/96cd28851e204d168383bf73d2a6bc7e.png
[21309232afc914b3cdaff88ff63ae6ff.png]: /images/20220828/fd18d7e615b44d6e9c3d0e6794c8999c.png
[91d0d74cba727f56a116253ce7d0567f.png]: /images/20220828/66afd9153a594617a7c566d3c3efdc32.png
[492da3404aef7faaf310e160b5e2b21e.png]: /images/20220828/d7ad445eb74643c8bc70df6a30587a27.png
[8c62d9a6c779436133517f9785fec7af.png]: /images/20220828/32350aa8d6974a0685decd7a3acf215b.png
[addbdc9b0d8f0ef74751a47cb128be7c.png]: /images/20220828/db6eb45a52f845febb0673e08abb3174.png
[4b2fed6793ffbf7343e977d7150c0a89.png]: /images/20220828/cd63c8aa585e461ea0b840211209a541.png
[Link 1]: https://link.zhihu.com/?target=https%3A//bbs.huaweicloud.com/blogs%3Futm_source%3Dzhihu%26utm_medium%3Dbbs-ex%26utm_campaign%3Dei%26utm_content%3Dcontent