Flink 自定义sink 写入 Clickhouse

Flink 自定义sink 写入 Clickhouse

Flink 自定义sink 写入 Clickhouse

• 添加依赖
• 基于 Flink 服务提交任务并执行时需要的依赖包
• 构建ClickhouseSink参数实例
• 构建自定义ClickhouseStoreSink

添加依赖

	ru.yandex.clickhouse
	clickhouse-jdbc
	0.2.4

基于 Flink 服务提交任务并执行时需要的依赖包

基于 flink 服务器提交任务前，先上传依赖包到 flink 的 lib 目录下；然后重启 flink 服务，使 jar 进行加载；否则会出现 ClassNoFoundException 或者 clickhouseNoClassDefFound：ErrCould not initialize class ru.yandex.clickhouse.ClickHouseDriver 的异常。

• clickhouse-jdbc-0.2.4.jar
• httpclient-4.5.2.jar
• httpcore-4.4.4.jar
• commons-logging-1.0.4.jar
• guava-19.0.jar

构建ClickhouseSink参数实例

public class ClickhouseSink implements Serializable {
	
	private static final long serialVersionUID = -4410041701538783205L;

	private String url;
	
	private String index;
	
	private String username;
	
	private String password;

	public String getUrl() {
		return url;
	}

	public String getIndex() {
		return index;
	}

	public String getUsername() {
		return username;
	}

	public String getPassword() {
		return password;
	}
	
	public ClickhouseSink(Object obj) {
		final JSONObject json = JSONObject.parseObject(obj.toString());
		this.url = json.getString("url");
		this.index = json.getString("index");
		this.username = json.getString("username");
		this.password = json.getString("password");
	}

}

构建自定义ClickhouseStoreSink

基于继承 RichSinkFunction< T > 抽象类实现自定义Sink，实现方法有三个：

• open()：构建sink节点时最先执行的方法，用于实现一些初始化动作。
• invoke()：执行节点时执行，用于实现具体业务逻辑。
• close()：关闭节点回收资源时执行，用于资源的回收。

import java.sql.Connection;
import java.sql.DriverManager;
import java.sql.Statement;
import java.util.Date;
import java.util.Map;

import org.apache.flink.configuration.Configuration;
import org.apache.flink.streaming.api.functions.sink.RichSinkFunction;

import com.alibaba.fastjson.JSONArray;
import com.alibaba.fastjson.JSONObject;
import com.ygsoft.dataprocess.vo.sink.ClickhouseSink;


public class ClickhouseStoreSink extends RichSinkFunction> {

	private static final long serialVersionUID = 5994842691528274078L;

	final private ClickhouseSink clickhouseSink;
	
	private Connection connection;
	
	public ClickhouseStoreSink(final ClickhouseSink clickhouseSink) {
		this.clickhouseSink = clickhouseSink;
	}
	
	
	@Override
	public void open(final Configuration parameters) throws Exception {
		super.open(parameters);
		Class.forName("ru.yandex.clickhouse.ClickHouseDriver");
		System.out.println(clickhouseSink.getUrl());
		connection = DriverManager.getConnection(clickhouseSink.getUrl(), clickhouseSink.getUsername(), clickhouseSink.getPassword());
		connection.setAutoCommit(false);
	}
	
	@Override
	public void invoke(Map value, Context context) throws Exception {
		Statement statement = connection.createStatement();
		final StringBuffer sqlCol = new StringBuffer("INSERT INTO ");
		sqlCol.append(clickhouseSink.getIndex() + " (deviceid, metric, timestamp, value, type, error, create_time) VALUES ");
		final JSONArray datas = JSONArray.parseArray(value.get("value"));
		final StringBuffer sqlVal = new StringBuffer();
		for(int i=0; i
欢迎分享，转载请注明来源：内存溢出
原文地址:https://www.54852.com/zaji/5690374.html