- FlinkX介绍
- 1 FlinkX的安装
- 2 FlinkX的简单使用
- 2.1 MySQLToHDFS
- 2.2 MySQLToHive
- 2.3 MySQLToHbase
- 2.4 MySQLToMySQL
介绍
FlinkX是在是袋鼠云内部广泛使用的基于flink的分布式离线和实时的数据同步框架，实现了多种异构数据源之间高效的数据迁移。
不同的数据源头被抽象成不同的Reader插件，不同的数据目标被抽象成不同的Writer插件。理论上，FlinkX框架可以支持任意数据源类型的数据同步工作。作为一套生态系统，每接入一套新数据源该新加入的数据源即可实现和现有的数据源互通。
FlinkX是一个基于Flink的批流统一的数据同步工具，既可以采集静态的数据，比如MySQL，HDFS等，也可以采集实时变化的数据，比如MySQL binlog，Kafka等
底层原理
1 FlinkX的安装在底层实现上，FlinkX依赖Flink，数据同步任务会被翻译成StreamGraph在Flink上执行，基本原理如下图:
安装unzip：yum install unzip
1、上传并解压
unzip flinkx-1.10.zip -d /usr/local/soft/
2、配置环境变量
vim /etc/profile #更新配置文件 source /etc/profile
3、给bin/flinkx这个文件加上执行权限
chmod a+x flinkx
4、修改配置文件，设置运行端口
vim flinkconf/flink-conf.yaml
## web服务端口，不指定的话会随机生成一个 rest.bind-port: 88882 FlinkX的简单使用 2.1 MySQLToHDFS
- 配置文件
{
"job": {
"content": [
{
"reader": {
"parameter": {
"username": "root",
"password": "123456",
"connection": [
{
"jdbcUrl": [
"jdbc:mysql://master:3306/student?characterEncoding=utf8"
],
"table": [
"student"
]
}
],
"column": [
"*"
],
"customSql": "",
"where": "clazz = '理科二班'",
"splitPk": "id",
"queryTimeOut": 1000,
"requestAccumulatorInterval": 2
},
"name": "mysqlreader"
},
"writer": {
"name": "hdfswriter",
"parameter": {
"path": "hdfs://master:9000/data/flinkx/student",
"defaultFS": "hdfs://master:9000",
"column": [
{
"name": "id",
"index": 0,
"type": "string"
},
{
"name": "name",
"index": 1,
"type": "string"
},
{
"name": "age",
"index": 2,
"type": "string"
},
{
"name": "gender",
"index": 3,
"type": "string"
},
{
"name": "clazz",
"index": 4,
"type": "string"
},
{
"name": "last_mod",
"index": 5,
"type": "string"
}
],
"fieldDelimiter": ",",
"fileType": "text",
"writeMode": "overwrite"
}
}
}
],
"setting": {
"restore": {
"isRestore": false
},
"speed": {
"channel": 1
}
}
}
}
- 启动任务
flinkx -mode local -job /usr/local/soft/flinkx-1.10/jsonConf/mysqlToHDFS.json -pluginRoot /usr/local/soft/flinkx-1.10/syncplugins/ -flinkconf /usr/local/soft/flinkx-1.10/flinkconf/
- 监听日志
flinkx 任务启动后，会在执行命令的目录下生成一个nohup.out文件
tail -f nohup.out
- 通过web界面查看任务运行情况
http://master:88882.2 MySQLToHive
- 配置文件
{
"job": {
"content": [
{
"reader": {
"parameter": {
"username": "root",
"password": "123456",
"connection": [
{
"jdbcUrl": [
"jdbc:mysql://master:3306/student?characterEncoding=utf8"
],
"table": [
"student"
]
}
],
"column": [
"*"
],
"customSql": "",
"where": "clazz = '文科二班'",
"splitPk": "id",
"queryTimeOut": 1000,
"requestAccumulatorInterval": 2
},
"name": "mysqlreader"
},
"writer": {
"name": "hivewriter",
"parameter": {
"jdbcUrl": "jdbc:hive2://master:10000/testflinkx",
"username": "",
"password": "",
"fileType": "text",
"fieldDelimiter": ",",
"writeMode": "overwrite",
"compress": "",
"charsetName": "UTF-8",
"maxFileSize": 1073741824,
"tablesColumn": "{"student":[{"key":"id","type":"string"},{"key":"name","type":"string"},{"key":"age","type":"string"}]}",
"defaultFS": "hdfs://master:9000"
}
}
}
],
"setting": {
"restore": {
"isRestore": false,
"isStream": false
},
"errorLimit": {},
"speed": {
"channel": 3
}
}
}
}
- 在hive中创建testflinkx数据库，并创建student分区表
create database testflinkx; use testflinkx; CREATE TABLE `student`( `id` string, `name` string, `age` string) PARTITIonED BY ( `pt` string) ROW FORMAT DELIMITED FIELDS TERMINATED BY ','
- 启动hiveserver2
# 第一种方式：hiveserver2 # 第二种方式:hive --service hiveserver2
- 启动任务
flinkx -mode local -job /usr/local/soft/flinkx-1.10/jsonConf/mysqlToHive.json -pluginRoot /usr/local/soft/flinkx-1.10/syncplugins/ -flinkconf /usr/local/soft/flinkx-1.10/flinkconf/
- 查看日志及运行情况同上
- 配置文件
{
"job": {
"content": [
{
"reader": {
"parameter": {
"username": "root",
"password": "123456",
"connection": [
{
"jdbcUrl": [
"jdbc:mysql://master:3306/student?characterEncoding=utf8"
],
"table": [
"score"
]
}
],
"column": [
"*"
],
"customSql": "",
"splitPk": "student_id",
"queryTimeOut": 1000,
"requestAccumulatorInterval": 2
},
"name": "mysqlreader"
},
"writer": {
"name": "hbasewriter",
"parameter": {
"hbaseConfig": {
"hbase.zookeeper.property.clientPort": "2181",
"hbase.rootdir": "hdfs://master:9000/hbase",
"hbase.cluster.distributed": "true",
"hbase.zookeeper.quorum": "master,node1,node2",
"zookeeper.znode.parent": "/hbase"
},
"table": "testFlinkx",
"rowkeyColumn": "$(cf1:student_id)_$(cf1:course_id)",
"column": [
{
"name": "cf1:student_id",
"type": "string"
},
{
"name": "cf1:course_id",
"type": "string"
},
{
"name": "cf1:score",
"type": "string"
}
]
}
}
}
],
"setting": {
"restore": {
"isRestore": false,
"isStream": false
},
"errorLimit": {},
"speed": {
"channel": 3
}
}
}
}
- 启动hbase 并创建testflinkx表
create 'testFlinkx','cf1'
- 启动任务
flinkx -mode local -job /usr/local/soft/flinkx-1.10/jsonConf/mysqlToHbase.json -pluginRoot /usr/local/soft/flinkx-1.10/syncplugins/ -flinkconf /usr/local/soft/flinkx-1.10/flinkconf/
- 查看日志及运行情况同上
- 配置文件
{
"job": {
"content": [
{
"reader": {
"name": "mysqlreader",
"parameter": {
"column": [
{
"name": "id",
"type": "int"
},
{
"name": "name",
"type": "string"
},
{
"name": "age",
"type": "int"
},
{
"name": "gender",
"type": "string"
},
{
"name": "clazz",
"type": "string"
}
],
"username": "root",
"password": "123456",
"connection": [
{
"jdbcUrl": [
"jdbc:mysql://master:3306/student?useSSL=false"
],
"table": [
"student"
]
}
]
}
},
"writer": {
"name": "mysqlwriter",
"parameter": {
"username": "root",
"password": "123456",
"connection": [
{
"jdbcUrl": "jdbc:mysql://master:3306/student?useSSL=false",
"table": [
"student2"
]
}
],
"writeMode": "insert",
"column": [
{
"name": "id",
"type": "int"
},
{
"name": "name",
"type": "string"
},
{
"name": "age",
"type": "int"
},
{
"name": "gender",
"type": "string"
},
{
"name": "clazz",
"type": "string"
}
]
}
}
}
],
"setting": {
"speed": {
"channel": 1,
"bytes": 0
}
}
}
}
欢迎分享，转载请注明来源：内存溢出
微信扫一扫
支付宝扫一扫
评论列表（0条）