一、Flume收集数据到Hbase中
1、自定义的Flume-HBase-Sink，主要完成的功能：
1、可以指定任意的分隔符
2、可以完成任意的列内容的分割
3、自定义行键
2、自定义FLume和HBase sink需要解决的问题
1、读到我们配置的列名有哪些
2、得要知道如何去拆分我们的数据
如果想要在这个类中获取自定义的配置项信息，那么该配置项必须要在Serializer.*的后面（*）来指定
二、具体实现步骤
1、下载flume源码包并解压：
下载地址：链接：https://pan.baidu.com/s/1IUGOTDRC2xZEqx9lHC90YQ 提取码：tom1
2、找到flume-ng-sinks文件夹并打开，用idea打开flume-ng-hbase-sink
3、在main/java下编写实现代码：
代码实现如下：
package org.apache.flume.sink.hbase;
import com.google.common.base.Charsets;
import org.apache.flume.Context;
import org.apache.flume.Event;
import org.apache.flume.FlumeException;
import org.apache.flume.conf.ComponentConfiguration;
import org.apache.hadoop.hbase.client.Increment;
import org.apache.hadoop.hbase.client.Put;
import org.apache.hadoop.hbase.client.Row;
import java.util.Arrays;
import java.util.LinkedList;
import java.util.List;
public class MySimpleHbaseEventSerializer implements HbaseEventSerializer {
private String rowPrefix;
private byte[] incrementRow;
private byte[] columnFamily;
private byte[] cf;
private String plCol;
private byte[] incCol;
private byte[] payload;
private String seperator;
public MySimpleHbaseEventSerializer() {
}
/**
* 获得配置文件中的配置项
* @param context
*/
@Override
public void configure(Context context) {
rowPrefix = context.getString("rowPrefix", "default");
incrementRow =
context.getString("incrementRow", "incRow").getBytes(Charsets.UTF_8);
String suffix = context.getString("suffix", "uuid");
//datetime,userid,searchname,retorder,cliorder,cliurl
String payloadColumn = context.getString("payloadColumn", "pCol");
String incColumn = context.getString("incrementColumn", "iCol");
String seperator = context.getString("seperator", ",");
String cf = context.getString("columnFamily", "cf1");
if (payloadColumn != null && !payloadColumn.isEmpty()) {
plCol = payloadColumn;//pCol
System.out.println("==============>payloadColumn: " + plCol);
}
if (incColumn != null && !incColumn.isEmpty()) {
incCol = incColumn.getBytes(Charsets.UTF_8);//iCol
}
if(seperator != null && !seperator.isEmpty()) {
this.seperator = seperator;
System.out.println("==============>seperator: " + seperator);
}
if(cf != null && !cf.isEmpty()) {
this.columnFamily = cf.getBytes(Charsets.UTF_8);
System.out.println("==============>columnFamily: " + cf);
}
}
@Override
public void configure(ComponentConfiguration conf) {
}
@Override
public void initialize(Event event, byte[] cf) {
this.payload = event.getBody();//读取到的数据
this.cf = cf;
}
/**
* 把数据由channel写入到hbase的操作就在这里面实现
* @return
* @throws FlumeException
*/
@Override
public List<Row> getActions() throws FlumeException {
List<Row> actions = new LinkedList<Row>();
if (plCol != null) {//pCol
byte[] rowKey = null;
try {
//数据在payload中，列信息在plCol中，这二者需要一一对应
String[] columns = this.plCol.split(",");
System.out.println("---------------->columns: " + Arrays.toString(columns));
String[] dataArray = new String(this.payload).split(this.seperator);
System.out.println("---------------->columns: " + Arrays.toString(dataArray));
if(columns == null || dataArray == null || columns.length != dataArray.length) {
return actions;
} else {
String userid = dataArray[1];
String time = dataArray[0];
Put put = null;
for (int i = 0; i < columns.length; i++) {
String column = columns[i];
String data = dataArray[i];
put = new Put(SimpleRowKeyGenerator.getUserIdAndTimeKey(time, userid));
put.addColumn(columnFamily, column.getBytes(), data.getBytes());
actions.add(put);
}
}
} catch (Exception e) {
throw new FlumeException("Could not get row key!", e);
}
}
return actions;
}
@Override
public List<Increment> getIncrements() {
List<Increment> incs = new LinkedList<Increment>();
if (incCol != null) {
Increment inc = new Increment(incrementRow);
inc.addColumn(cf, incCol, 1);
incs.add(inc);
}
return incs;
}
@Override
public void close() {
}
}
4、打成jar包替换linux下的jar包
5、编写Flume配置文件：
主要作用是文件中的新增内容，将数据打入到HBase中
注意：Flume agent的运行，主要就是配置source channel sink
下面的a1就是agent的代号，source叫r1 channel叫c1 sink叫k1
a1.sources = r1
a1.sinks = k1
a1.channels = c1
#对于source的配置描述 监听目录中的新增数据
a1.sources.r1.type = exec
a1.sources.r1.command = tail -F /home/bigdata/data/projects/news/data/news-logs.log
#对于sink的配置描述 使用hbase做数据的消费
a1.sinks.k1.type = asynchbase
a1.sinks.k1.table = news_log_1805
a1.sinks.k1.columnFamily = cf
a1.sinks.k1.serializer = org.apache.flume.sink.hbase.MyAsyncHbaseEventSerializer
## 将flume采集到的数据，写入到hbase表中对应的列
a1.sinks.k1.serializer.payloadColumn = datetime,userid,searchname,retorder,cliorder,cliurl
a1.sinks.k1.serializer.separator = \\t
#对于channel的配置描述 使用内存缓冲区域做数据的临时缓存,文件
a1.channels.c1.type = file
a1.channels.c1.checkpointDir = /home/bigdata/data/projects/news/checkpoint
a1.channels.c1.dataDirs = /home/bigdata/data/projects/news/channel
a1.channels.c1.capacity = 100000
a1.channels.c1.transactionCapacity = 10000
#通过channel c1将source r1和sink k1关联起来
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c16、后台运行flume
nohup bin/flume-ng agent -n a1 -c conf -f conf/flume-hbase-sink.conf >/dev/null 2>&1 &