flume自定义sink写入es

flume官方没有提供写入elasticsearch的sink
所以只有自定义sink
这里采用的是flume1.9.0
elasticsearch采用的是7.17.0

1.首先创建maven项目
2.引入依赖

    <dependency>
        <groupId>co.elastic.clients</groupId>
        <artifactId>elasticsearch-java</artifactId>
        <version> ${elasticsearch.version} </version>
    </dependency>

    <dependency>
        <groupId>org.projectlombok</groupId>
        <artifactId>lombok</artifactId>
        <version>1.16.22</version>
        <scope>compile</scope>
    </dependency>



    <dependency>
        <groupId> com.fasterxml.jackson.core </groupId>
        <artifactId> jackson-databind </artifactId>
        <version>2.12.3</version>
    </dependency>

    <dependency>
        <groupId>org.apache.flume</groupId>
        <artifactId>flume-ng-core</artifactId>
        <version>1.9.0</version>
    </dependency>

3.编写：写入es数据的类
package com._3men.es;

import co.elastic.clients.elasticsearch.ElasticsearchClient;
import co.elastic.clients.elasticsearch.core.CreateRequest;
import co.elastic.clients.elasticsearch.core.CreateResponse;
import co.elastic.clients.json.jackson.JacksonJsonpMapper;
import co.elastic.clients.transport.ElasticsearchTransport;
import co.elastic.clients.transport.rest_client.RestClientTransport;
import com._3men.ToolUtil;
import lombok.extern.slf4j.Slf4j;
import org.apache.http.HttpHost;
import org.elasticsearch.client.RestClient;

import java.io.IOException;

/**
*对es操作简单进行封装
*/
@Slf4j
public class ElasticsearchTool{

private  ElasticsearchClient client;

private  String hosts;

private  String defaultIndex;

public ElasticsearchTool(String hosts,String defaultIndex){
    this.hosts=hosts;
    this.defaultIndex=defaultIndex;

    RestClient restClient = RestClient.builder(getHttpHost(hosts)).build();
    ElasticsearchTransport transport = new RestClientTransport(restClient, new JacksonJsonpMapper());
    client = new ElasticsearchClient(transport);
  
}


private static HttpHost[] getHttpHost(String hosts){
    String[] hostsSplit = hosts.split(",");
    HttpHost[] hostArr=new HttpHost[hostsSplit.length];
    for (int i = 0; i < hostsSplit.length; i++) {
        String[] addressSplit = hostsSplit[i].split(":");
        hostArr[i]=new HttpHost(addressSplit[0],Integer.valueOf(addressSplit[1]));
    }
    return hostArr;
}


public   <T>  CreateResponse createDocument(T t,String id,String index) throws IOException {
    return client.create(new CreateRequest.Builder<T>().document(t).index(index).id(id).build());
}

public   <T>  CreateResponse createDocument(T t,String id) throws IOException {
   return createDocument(t,id,defaultIndex);
}

public   <T>  CreateResponse createDocument(T t) throws IOException {
    String uuid = ToolUtil.uuid();
    CreateResponse response = createDocument(t, uuid, defaultIndex);
    return response;

}


public  void stop(){
    try {
        client._transport().close();
    } catch (IOException e) {
        e.printStackTrace();
    }
}

}

4.自定义sink
package com._3men.sink;

import co.elastic.clients.elasticsearch.core.CreateResponse;
import com._3men.bean.Journal;
import com._3men.es.ElasticsearchTool;
import lombok.extern.slf4j.Slf4j;
import org.apache.flume.*;
import org.apache.flume.conf.Configurable;
import org.apache.flume.sink.AbstractSink;

@Slf4j
public class ElasticsearchSink extends AbstractSink implements Configurable {

private ElasticsearchTool tool;

@Override
public void configure(Context context) {
    String hosts = context.getString("es.cluster");
    String index = context.getString("es.index");
    tool=new ElasticsearchTool(hosts,index);
}

@Override
public synchronized void stop() {
    super.stop();
    tool.stop();
}

@Override
public Status process()  {
    Status status = null;
    Channel ch = getChannel();
    Transaction txn = ch.getTransaction();

    txn.begin();
    try {
        Event event = ch.take();
        //1.解析event 然后拿到需要插入的数据， 这里有可能返回为空
        if (event==null||event.getHeaders()==null){
            log.info("event或者 event header为空"+event);
        }else {
                //数据转换  自己根据业务实现
                Object o=//自定义的转换方法
                tool.createDocument(o);
        }
        txn.commit();
        status = Status.READY;
    } catch (Throwable t) {
        txn.rollback();
        status = Status.BACKOFF;
        if (t instanceof Error) {
            throw (Error)t;
        }
        log.error("异常了：{}",t);
    }finally {
        txn.close();
    }
    return status;
}

}

4.将项目打成jar包

5.由于maven打包不会将依赖打进去，所以需要将依赖的jar包加入到flume的lib目录下,由于这里会有一些jar冲突，这里给出需要删除的jar包和添加jar包
添加：

删除原来flumelib中的

6.导出项目依赖jar包

配置文件编写
a1.sources = r1
a1.sinks = k1
a1.channels = c1

1.定义source 这里采用 TAILDIR souce

a1.sources.r1.type = TAILDIR
a1.sources.r1.positionFile = /home/apache-flume-1.9.0/taildir_position.json
a1.sources.r1.filegroups = f1 f2
a1.sources.r1.fileHeader = true
a1.sources.ri.maxBatchCount = 100

a1.sources.r1.filegroups.f1 = /home/log/smso-crm-api/.*.log
a1.sources.r1.headers.f1.env = 测试环境
a1.sources.r1.headers.f1.app = 开票系统
a1.sources.r1.headers.f1.service = 开票系统

a1.sources.r1.filegroups.f2 = /home/log/data-service/.*.log
a1.sources.r1.headers.f2.env = 测试环境
a1.sources.r1.headers.f2.app = 数据迁移中心
a1.sources.r1.headers.f2.service = 数据迁移中心

#2.定义管道 这里用内存管道
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 200

#3.定义sink
a1.sinks.k1.es.cluster=192.168.247.100:19200
a1.sinks.k1.es.index=journals
a1.sinks.k1.type = com._3men.sink.ElasticsearchSink

#sink和source 绑定管道
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

#最后启动