flume的配置详解

Flume：
=====================
Flume是一种分布式的、可靠的、可用的服务，可以有效地收集、聚合和移动大量的日志数据。
它有一个基于流数据的简单而灵活的体系结构。
它具有健壮性和容错能力，具有可调的可靠性机制和许多故障转移和恢复机制。
它使用一个简单的可扩展数据模型，允许在线分析应用程序。

source：源
对channel而言，相当于生产者，通过接收各种格式数据发送给channel进行传输

channel：通道
相当于数据缓冲区，接收source数据发送给sink

sink：沉槽
对channel而言，相当于消费者，通过接收channel数据通过指定数据类型发送到指定位置

Event：
===============
flume传输基本单位：
head + body

flume安装：
================
1、解压
2、符号链接
3、配置环境变量并使其生效
4、修改配置文件
1)重命名flume-env.ps1.template为flume-env.ps1
2)重命名flume-env.sh.template为flume-env.sh
3)修改flume-env.sh，配置jdk目录,添加
export JAVA_HOME=/soft/jdk

5、flume 查看版本
flume-ng version

flume使用：
=========================
//flume可以将配置文件写在zk上

//flume运行命令
flume-ng agent -n a1 -f xxx.conf /flume-ng agent -n xx -f xxx.conf

agent: a1
source: s1
channel:c1
sink: n1

使用方法：
1、编写配置文件r_nc.conf
# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = netcat
a1.sources.r1.bind = localhost
a1.sources.r1.port = 8888

# 配置sink
a1.sinks.k1.type = logger

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

2、启动flume，指定配置文件
flume-ng agent -n a1 -f r_nc.conf

3、启动另一个会话，进行测试
nc localhost 8888

//用户手册
http://flume.apache.org/FlumeUserGuide.html

后台运行程序：
=============================================

ctrl + z :将程序放在后台运行 =====> [1]+ Stopped flume-ng agent -n a1 -f r_nc.conf

通过 bg %1 的方式将程序后台运行

通过jobs查看后台任务

通过 fg %1 的方式将程序放在前台运行

flume:
海量日志数据的收集、聚合和移动

flume-ng agent -n a1 -f xxx.conf

source
相对于channel是生产者 //netcat
channel
类似于缓冲区 //memory
sink
相对于channel是消费者 //logger

Event:
header + body
k v data

source:
============================================
1、序列(seq)源：多用作测试
# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = seq
# 总共发送的事件个数
a1.sources.r1.totalEvents = 1000

# 配置sink
a1.sinks.k1.type = logger

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

2、压力(stress)源:多用作负载测试
# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = org.apache.flume.source.StressSource
# 单个事件大小,单位：byte
a1.sources.r1.size = 10240
# 事件总数
a1.sources.r1.maxTotalEvents = 1000000

# 配置sink
a1.sinks.k1.type = logger

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

3、滚动目录(Spooldir)源：监听指定目录新文件产生，并将新文件数据作为event发送
# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = spooldir
# 设置监听目录
a1.sources.r1.spoolDir = /home/centos/spooldir

# 通过以下配置指定消费完成后文件后缀
#a1.sources.r1.fileSuffix = .COMPLETED

# 配置sink
a1.sinks.k1.type = logger

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

4、exec源 //通过执行linux命令产生新数据
//典型应用 tail -F (监听一个文件,文件增长的时候，输出追加数据)
//不能保证数据完整性，很可能丢失数据

# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = exec
# 配置linux命令
a1.sources.r1.command = tail -F /home/centos/readme.txt

# 配置sink
a1.sinks.k1.type = logger

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

5、Taildir源 //监控目录下文件
//文件类型可通过正则指定
//有容灾机制

# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = TAILDIR
# 设置source组 可设置多个
a1.sources.r1.filegroups = f1
# 设置组员的监控目录和监控文件类型,使用正则表示,只能监控文件
a1.sources.r1.filegroups.f1 = /home/centos/taildir/.*

# 设置定位文件的位置
# a1.sources.r1.positionFile ~/.flume/taildir_position.json

# 配置sink
a1.sinks.k1.type = logger

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

sink：
====================================
1、fileSink //多用作数据收集
# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = netcat
a1.sources.r1.bind = localhost
a1.sources.r1.port = 8888

# 配置sink
a1.sinks.k1.type = file_roll
# 配置目标文件夹
a1.sinks.k1.sink.directory = /home/centos/file
# 设置滚动间隔，默认30s，设为0则不滚动，成为单个文件
a1.sinks.k1.sink.rollInterval = 0

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

2、hdfsSink //默认以seqFile格式写入
//k：LongWritable
//v: BytesWritable
//
# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = netcat
a1.sources.r1.bind = localhost
a1.sources.r1.port = 8888

# 配置sink
a1.sinks.k1.type = hdfs
# 配置目标文件夹
a1.sinks.k1.hdfs.path = /flume/events/%y-%m-%d/
# 配置文件前缀
a1.sinks.k1.hdfs.filePrefix = events-
# 滚动间隔，秒
a1.sinks.k1.hdfs.rollInterval = 0
# 触发滚动文件大小，byte
a1.sinks.k1.hdfs.rollSize = 1024
# 配置使用本地时间戳
a1.sinks.k1.hdfs.useLocalTimeStamp = true
# 配置输出文件类型，默认SequenceFile
# DataStream文本格式,不能设置压缩编解码器
# CompressedStream压缩文本格式，需要设置编解码器
a1.sinks.k1.hdfs.fileType = DataStream

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

3、hiveSink： //hiveserver帮助：hive --service help
//1、hive --service metastore 启动hive的metastore服务，metastore地址：thrift://localhost:9083
//2、将hcatalog的依赖放在/hive/lib下，cp hive-hcatalog* /soft/hive/lib (位置/soft/hive/hcatalog/share/hcatalog)
//3、创建hive事务表
//SET hive.support.concurrency=true;
SET hive.enforce.bucketing=true;
SET hive.exec.dynamic.partition.mode=nonstrict;
SET hive.txn.manager=org.apache.hadoop.hive.ql.lockmgr.DbTxnManager;
SET hive.compactor.initiator.on=true;
SET hive.compactor.worker.threads=1;

//create table myhive.weblogs(id int, name string, age int)
clustered by(id) into 2 buckets
row format delimited
fields terminated by '\t'
stored as orc
tblproperties('transactional'='true');

# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = netcat
a1.sources.r1.bind = localhost
a1.sources.r1.port = 8888

# 配置sink
a1.sinks.k1.type = hive
a1.sinks.k1.hive.metastore = thrift://127.0.0.1:9083
a1.sinks.k1.hive.database = myhive
a1.sinks.k1.hive.table = weblogs
a1.sinks.k1.useLocalTimeStamp = true
#输入格式，DELIMITED和json
#DELIMITED 普通文本
#json json文件
a1.sinks.k1.serializer = DELIMITED
#输入字段分隔符,双引号
a1.sinks.k1.serializer.delimiter = ","
#输出字段分隔符,单引号
a1.sinks.k1.serializer.serdeSeparator = '\t'
#字段名称，","分隔，不能有空格
a1.sinks.k1.serializer.fieldnames =id,name,age

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

4、hbaseSink //SimpleHbaseEventSerializer将rowKey和col设置了默认值，不能自定义
//RegexHbaseEventSerializer可以手动指定rowKey和col字段名称

# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = netcat
a1.sources.r1.bind = localhost
a1.sources.r1.port = 8888

# 配置sink
a1.sinks.k1.type = hbase
a1.sinks.k1.table = flume_hbase
a1.sinks.k1.columnFamily = f1
a1.sinks.k1.serializer = org.apache.flume.sink.hbase.RegexHbaseEventSerializer

# 配置col正则手动指定
# rowKeyIndex手动指定rowKey，索引以0开头
a1.sinks.k1.serializer.colNames = ROW_KEY,name,age
a1.sinks.k1.serializer.regex = (.*),(.*),(.*)
a1.sinks.k1.serializer.rowKeyIndex=0

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

5、asynchbaseSink //异步hbaseSink
//异步机制，写入速度快
# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = netcat
a1.sources.r1.bind = localhost
a1.sources.r1.port = 8888

# 配置sink
a1.sinks.k1.type = asynchbase
a1.sinks.k1.table = flume_hbase
a1.sinks.k1.columnFamily = f1
a1.sinks.k1.serializer = org.apache.flume.sink.hbase.SimpleAsyncHbaseEventSerializer

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

channel：缓冲区
=====================================
1、memorychannel
a1.channels.c1.type = memory
# 缓冲区中存留的最大event个数
a1.channels.c1.capacity = 1000
# channel从source中每个事务提取的最大event数
# channel发送给sink每个事务发送的最大event数
a1.channels.c1.transactionCapacity = 100

2、fileChannel： //检查点和数据存储在默认位置时，当多个channel同时开启
//会导致文件冲突，引发其他channel会崩溃

# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = netcat
a1.sources.r1.bind = localhost
a1.sources.r1.port = 8888

# 配置sink
a1.sinks.k1.type = logger

# 配置channel
a1.channels = c1
a1.channels.c1.type = file
a1.channels.c1.checkpointDir = /home/centos/flume/checkpoint
a1.channels.c1.dataDirs = /home/centos/flume/data

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

memoryChannel：快速，但是当设备断电，数据会丢失

FileChannel： 速度较慢，即使设备断电，数据也不会丢失

Avro
===============================================
source
# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = avro
a1.sources.r1.bind = 0.0.0.0
a1.sources.r1.port = 4444

# 配置sink
a1.sinks.k1.type = logger

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

***********************************************************************************************
*启动avro客户端，发送数据： *
* flume-ng avro-client -H localhost -p 4444 -R ~/avro/header.txt -F ~/avro/user0.txt *
* 指定ip 指定端口 指定header文件 指定数据文件 *
***********************************************************************************************

sink
# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = TAILDIR
a1.sources.r1.filegroups = f1
a1.sources.r1.filegroups.f1 = /home/centos/taildir/.*

# 配置sink
a1.sinks.k1.type = avro
a1.sinks.k1.bind = 192.168.23.101
a1.sinks.k1.port = 4444

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

Flume跃点：
=====================================
1、将s101的flume发送到其他节点
xsync.sh /soft/flume
xsync.sh /soft/apache-flume-1.8.0-bin/

2、切换到root用户，分发环境变量文件
su root
xsync.sh /etc/profile
exit

3、配置文件
1)配置s101 //hop.conf
设置source:avro
设置sink： hdfs

# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = avro
a1.sources.r1.bind = 0.0.0.0
a1.sources.r1.port = 4444

# 配置sink
a1.sinks.k1.type = hdfs
a1.sinks.k1.hdfs.path = /flume/hop/%y-%m-%d/
a1.sinks.k1.hdfs.filePrefix = events-
a1.sinks.k1.hdfs.rollInterval = 0
a1.sinks.k1.hdfs.rollSize = 1024
a1.sinks.k1.hdfs.useLocalTimeStamp = true
a1.sinks.k1.hdfs.fileType = DataStream

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

2)配置s102-s104 //hop2.conf
设置source:taildir
设置sink： avro

# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = TAILDIR
a1.sources.r1.filegroups = f1
a1.sources.r1.filegroups.f1 = /home/centos/taildir/.*

# 配置sink
a1.sinks.k1.type = avro
a1.sinks.k1.hostname = 192.168.23.101
a1.sinks.k1.port = 4444

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

4、在s102-s104创建~/taildir文件夹
xcall.sh "mkdir ~/taildir"

5、启动s101的flume
flume-ng agent -n a1 -f /soft/flume/conf/hop.conf

6、分别启动s102-s104的flume,并将其放在后台运行
flume-ng agent -n a1 -f /soft/flume/conf/hop2.conf &

7、进行测试，分别在s102-s104的taildir中创建数据，观察hdfs数据情况
s102]$ echo 102 > taildir/1.txt
s103]$ echo 103 > taildir/1.txt
s104]$ echo 104 > taildir/1.txt

interceptor：拦截器
==================================
是source端组件：负责修改或删除event
每个source可以配置多个拦截器 ===> interceptorChain

1、Timestamp Interceptor //时间戳拦截器 + header

# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = netcat
a1.sources.r1.bind = localhost
a1.sources.r1.port = 8888
# 给拦截器起名
a1.sources.r1.interceptors = i1
# 指定拦截器类型
a1.sources.r1.interceptors.i1.type = timestamp

# 配置sink
a1.sinks.k1.type = logger

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

2、Static Interceptor //静态拦截器 + header

3、Host Interceptor //主机拦截器 + header

4、设置拦截器链：

# 将agent组件起名
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 配置source
a1.sources.r1.type = netcat
a1.sources.r1.bind = localhost
a1.sources.r1.port = 8888

a1.sources.r1.interceptors = i1 i2 i3
a1.sources.r1.interceptors.i1.type = timestamp
a1.sources.r1.interceptors.i2.type = host
a1.sources.r1.interceptors.i3.type = static
a1.sources.r1.interceptors.i3.key = location
a1.sources.r1.interceptors.i3.value = NEW_YORK

# 配置sink
a1.sinks.k1.type = logger

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

channel selector：通道挑选器
====================================
是source端组件：负责将event发送到指定的channel，相当于分区

当一个source设置多个channel时，默认以副本形式向每个channel发送一个event拷贝

1、replication副本通道挑选器 //默认挑选器，source将所有channel发送event副本
//设置source x 1, channel x 3, sink x 3
// nc memory file

# 将agent组件起名
a1.sources = r1
a1.sinks = k1 k2 k3
a1.channels = c1 c2 c3

# 配置source
a1.sources.r1.type = netcat
a1.sources.r1.bind = localhost
a1.sources.r1.port = 8888
a1.sources.r1.selector.type = replicating

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

a1.channels.c2.type = memory
a1.channels.c2.capacity = 1000
a1.channels.c2.transactionCapacity = 100

a1.channels.c3.type = memory
a1.channels.c3.capacity = 1000
a1.channels.c3.transactionCapacity = 100

# 配置sink
a1.sinks.k1.type = file_roll
a1.sinks.k1.sink.directory = /home/centos/file1
a1.sinks.k1.sink.rollInterval = 0

a1.sinks.k2.type = file_roll
a1.sinks.k2.sink.directory = /home/centos/file2
a1.sinks.k2.sink.rollInterval = 0

a1.sinks.k3.type = file_roll
a1.sinks.k3.sink.directory = /home/centos/file3
a1.sinks.k3.sink.rollInterval = 0

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1 c2 c3
a1.sinks.k1.channel = c1
a1.sinks.k2.channel = c2
a1.sinks.k3.channel = c3

2、Multiplexing 多路复用通道挑选器 //选择avro源发送文件

# 将agent组件起名
a1.sources = r1
a1.sinks = k1 k2 k3
a1.channels = c1 c2 c3

# 配置source
a1.sources.r1.type = avro
a1.sources.r1.bind = 0.0.0.0
a1.sources.r1.port = 4444
# 配置通道挑选器
a1.sources.r1.selector.type = multiplexing
a1.sources.r1.selector.header = country
a1.sources.r1.selector.mapping.CN = c1
a1.sources.r1.selector.mapping.US = c2
a1.sources.r1.selector.default = c3

# 配置channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

a1.channels.c2.type = memory
a1.channels.c2.capacity = 1000
a1.channels.c2.transactionCapacity = 100

a1.channels.c3.type = memory
a1.channels.c3.capacity = 1000
a1.channels.c3.transactionCapacity = 100

# 配置sink
a1.sinks.k1.type = file_roll
a1.sinks.k1.sink.directory = /home/centos/file1
a1.sinks.k1.sink.rollInterval = 0

a1.sinks.k2.type = file_roll
a1.sinks.k2.sink.directory = /home/centos/file2
a1.sinks.k2.sink.rollInterval = 0

a1.sinks.k3.type = file_roll
a1.sinks.k3.sink.directory = /home/centos/file3
a1.sinks.k3.sink.rollInterval = 0

# 绑定channel-source, channel-sink
a1.sources.r1.channels = c1 c2 c3
a1.sinks.k1.channel = c1
a1.sinks.k2.channel = c2
a1.sinks.k3.channel = c3

1、创建file1 file2 file3文件夹，家目录
mkdir file1 file2 file3

2、创建文件夹country，并放入头文件和数据
创建头文件CN.txt、US.txt、OTHER.txt
CN.txt ===> country CN
US.txt ===> country US
OTHER.txt ===> country OTHER

创建数据 1.txt
1.txt ====> helloworld

3、运行flume
flume-ng agent -n a1 -f /soft/flume/selector_multi.conf

4、运行Avro客户端
flume-ng avro-client -H localhost -p 4444 -R ~/country/US.txt -F ~/country/1.txt ===> 查看file2
flume-ng avro-client -H localhost -p 4444 -R ~/country/CN.txt -F ~/country/1.txt ===> 查看file1
flume-ng avro-client -H localhost -p 4444 -R ~/country/OTHER.txt -F ~/country/1.txt ===> 查看file3

sinkProcessor
=================================
sink Runner 运行一个 sink Group

sink Group 是由一个或多个 sink 构成

sink Runner 告诉 sink Group 处理下一批 event

sink Group 含有一个 sink Processor , 负责指定一个 sink 来处理这批数据

2、failover 容灾 //将所有sink设置一个优先级
//数量越大，优先级越高
//当数据传入时，优先级最高的sink负责处理
//当sink挂掉，次高优先级的sink被激活，继续处理数据
//channel和sink必须一对一

a1.sources = r1
a1.sinks = s1 s2 s3
a1.channels = c1 c2 c3

# Describe/configure the source
a1.sources.r1.type = seq

a1.sinkgroups = g1
a1.sinkgroups.g1.sinks = s1 s2 s3
a1.sinkgroups.g1.processor.type = failover
a1.sinkgroups.g1.processor.priority.s1 = 5
a1.sinkgroups.g1.processor.priority.s2 = 10
a1.sinkgroups.g1.processor.priority.s3 = 15
a1.sinkgroups.g1.processor.maxpenalty = 10000

# Describe the sink
a1.sinks.s1.type = file_roll
a1.sinks.s1.sink.directory = /home/centos/file1
a1.sinks.s2.type = file_roll
a1.sinks.s2.sink.directory = /home/centos/file2
a1.sinks.s3.type = file_roll
a1.sinks.s3.sink.directory = /home/centos/file3

# Use a channel which buffers events in memory
a1.channels.c1.type = memory
a1.channels.c2.type = memory
a1.channels.c3.type = memory

# Bind the source and sink to the channel
a1.sources.r1.channels = c1 c2 c3
a1.sinks.s1.channel = c1
a1.sinks.s2.channel = c2
a1.sinks.s3.channel = c3

Event事件是由Source端封装输入数据的字节数组得来的
Event event = EventBuilder.withBody(body);

Sink中的process方法返回两种状态：
1、READY //一个或多个event成功分发
2、BACKOFF //channel中没有数据提供给sink

flume中事务的生命周期：

tx.begin() //开启事务，之后执行操作
tx.commit() //提交事务，操作完成后由此提交
tx.rollback() //回滚事务，出现异常可以采取回滚措施
tx.close() //关闭事务，最后一定要关闭事务