【Hadoop】10、Flume组件
Flume组件安装配置
1、下载和解压 Flume
# 传Flume安装包
[root@master ~]# cd /opt/software/
[root@master software]# ls
apache-flume-1.6.0-bin.tar.gz hadoop-2.7.1.tar.gz jdk-8u152-linux-x64.tar.gz mysql-5.7.18.zip sqoop-1.4.7.bin__hadoop-2.6.0.tar.gz
apache-hive-2.0.0-bin.tar.gz hbase-1.2.1-bin.tar.gz mysql-5.7.18 mysql-connector-java-5.1.46.jar zookeeper-3.4.8.tar.gz
# 使用root用户解压Flume到“/usr/local/src”路径
[root@master software]# tar xf /opt/software/apache-flume-1.6.0-bin.tar.gz -C /usr/local/src/
# 修改Flume安装路径文件夹名称
[root@master software]# cd /usr/local/src
[root@master src]# mv apache-flume-1.6.0-bin flume
# 修改文件夹归属用户和归属组为hadoop用户和hadoop组
[root@master src]# chown -R hadoop.hadoop /usr/local/src/
2、Flume 组件部署
# 编辑系统环境变量配置文件
[root@master src]# vi /etc/profile.d/flume.sh
添加：
export FLUME_HOME=/usr/local/src/flume
export PATH=${FLUME_HOME}/bin:$PATH
# 切换hadoop用户
[root@master src]# su - hadoop
# 查看是否成功
[hadoop@master ~]$ echo $PATH
/usr/local/src/zookeeper/bin:/usr/local/src/sqoop/bin:/usr/local/src/hbase/bin:/usr/local/src/jdk/bin:/usr/local/src/hadoop/bin:/usr/local/src/hadoop/sbin:/usr/local/src/flume/bin:/usr/local/bin:/bin:/usr/bin:/usr/local/sbin:/usr/sbin:/usr/local/src/hive/bin:/home/hadoop/.local/bin:/home/hadoop/bin
# 修改hbase-env.sh文件
[hadoop@master ~]$ vim /usr/local/src/hbase/conf/hbase-env.sh
#export HBASE_CLASSPATH=/usr/local/src/hadoop/etc/hadoop/ 注释掉这一行的内容
# 拷贝 flume-env.sh.template 文件
[hadoop@master ~]$ cd /usr/local/src/flume/conf
[hadoop@master conf]$ cp flume-env.sh.template flume-env.sh
# 修改并配置 flume-env.sh 文件
[hadoop@master conf]$ vi flume-env.sh
修改：
export JAVA_HOME=/usr/local/src/jdk
# 启动hadoop
[hadoop@master conf]$ start-all.sh
This script is Deprecated. Instead use start-dfs.sh and start-yarn.sh
Starting namenodes on [master]
hadoop@master's password:
master: namenode running as process 50448. Stop it first.
192.168.100.30: datanode running as process 43460. Stop it first.
192.168.100.20: datanode running as process 46094. Stop it first.
Starting secondary namenodes [0.0.0.0]
hadoop@0.0.0.0's password:
0.0.0.0: secondarynamenode running as process 50670. Stop it first.
starting yarn daemons
resourcemanager running as process 50836. Stop it first.
192.168.100.30: nodemanager running as process 43584. Stop it first.
192.168.100.20: nodemanager running as process 46228. Stop it first.
# 执行以上命令后要确保master上有NameNode、SecondaryNameNode、ResourceManager进程，在slave节点上要能看到DataNode、NodeManager进程
# master节点查看
[hadoop@master conf]$ jps
50448 NameNode
50836 ResourceManager
47502 QuorumPeerMain
50670 SecondaryNameNode
55855 Jps
# slave1节点查看
[root@slave1 ~]# su - hadoop
[hadoop@slave1 ~]$ jps
2070 DataNode
2364 Jps
2191 NodeManager
[hadoop@slave1 ~]$
# slave2节点查看
[root@slave2 ~]# su - hadoop
[hadoop@slave2 ~]$ jps
2166 NodeManager
2055 DataNode
2345 Jps
[hadoop@slave2 ~]$
# 验证安装是否成功
[hadoop@master conf]$ flume-ng version
Flume 1.6.0
Source code repository: https://git-wip-us.apache.org/repos/asf/flume.git
Revision: 2561a23240a71ba20bf288c7c2cda88f443c2080
Compiled by hshreedharan on Mon May 11 11:15:44 PDT 2015
From source with checksum b29e416802ce9ece3269d34233baf43f
# 若能够正常查询 Flume 组件版本为1.6.0，则表示安装成功。
3、使用 Flume 发送和接受信息
# 在 Flume 安装目录中创建 simple-hdfs-flume.conf 文件
[hadoop@master conf]$ cd /usr/local/src/flume
[hadoop@master flume]$ vi simple-hdfs-flume.conf
a1.sources=r1
a1.sinks=k1
a1.channels=c1
a1.sources.r1.type=spooldir
a1.sources.r1.spoolDir=/usr/local/src/hadoop/logs
a1.sources.r1.fileHeader=true
a1.sources.r1.deserializer.maxLineLength=30000
a1.sinks.k1.type=hdfs
a1.sinks.k1.hdfs.path=hdfs://master:9000/tmp/flume
a1.sinks.k1.hdfs.rollsize=1024000
a1.sinks.k1.hdfs.rollCount=0
a1.sinks.k1.hdfs.rollInterval=900
a1.sinks.k1.hdfs.useLocalTimeStamp=true
a1.channels.c1.type=file
a1.channels.c1.capacity=10000
a1.channels.c1.transactionCapacity=1000
a1.sources.r1.channels=c1
a1.sinks.k1.channel=c1
# 删除/tmp
[hadoop@master flume]$ hdfs dfs -rm -r /tmp
22/05/08 22:16:44 INFO fs.TrashPolicyDefault: Namenode trash configuration: Deletion interval = 0 minutes, Emptier interval = 0 minutes.
Deleted /tmp
# 创建/tmp/flume
[hadoop@master flume]$ hdfs dfs -mkdir -p /tmp/flume
# 查看文件
[hadoop@master flume]$ hdfs dfs -ls /
Found 5 items
drwxr-xr-x - hadoop supergroup 0 2022-04-29 15:26 /hbase
drwxr-xr-x - hadoop supergroup 0 2022-04-29 11:59 /input
drwxr-xr-x - hadoop supergroup 0 2022-04-29 12:00 /output
drwxr-xr-x - hadoop supergroup 0 2022-05-08 22:16 /tmp
drwxr-xr-x - hadoop supergroup 0 2022-04-29 16:49 /user
# 使用 flume-ng agent 命令加载 simple-hdfs-flume.conf 配置信息，启动 flume 传输数据
[hadoop@master flume]$ flume-ng agent --conf-file simple-hdfs-flume.conf --name a1
22/05/08 22:17:56 INFO hdfs.BucketWriter: Creating hdfs://master:9000/tmp/flume/FlumeData.1652019471497.tmp
22/05/08 22:17:56 INFO hdfs.BucketWriter: Closing hdfs://master:9000/tmp/flume/FlumeData.1652019471497.tmp
22/05/08 22:17:56 INFO hdfs.BucketWriter: Renaming hdfs://master:9000/tmp/flume/FlumeData.1652019471497.tmp to hdfs://master:9000/tmp/flume/FlumeData.1652019471497
22/05/08 22:17:56 INFO hdfs.BucketWriter: Creating hdfs://master:9000/tmp/flume/FlumeData.1652019471498.tmp
22/05/08 22:17:56 INFO hdfs.BucketWriter: Closing hdfs://master:9000/tmp/flume/FlumeData.1652019471498.tmp
22/05/08 22:17:56 INFO hdfs.BucketWriter: Renaming hdfs://master:9000/tmp/flume/FlumeData.1652019471498.tmp to hdfs://master:9000/tmp/flume/FlumeData.1652019471498
22/05/08 22:17:56 INFO hdfs.BucketWriter: Creating hdfs://master:9000/tmp/flume/FlumeData.1652019471499.tmp
22/05/08 22:17:56 INFO hdfs.BucketWriter: Closing hdfs://master:9000/tmp/flume/FlumeData.1652019471499.tmp
22/05/08 22:17:56 INFO hdfs.BucketWriter: Renaming hdfs://master:9000/tmp/flume/FlumeData.1652019471499.tmp to hdfs://master:9000/tmp/flume/FlumeData.1652019471499
22/05/08 22:17:56 INFO hdfs.BucketWriter: Creating hdfs://master:9000/tmp/flume/FlumeData.1652019471500.tmp
22/05/08 22:17:57 INFO hdfs.BucketWriter: Closing hdfs://master:9000/tmp/flume/FlumeData.1652019471500.tmp
22/05/08 22:17:57 INFO hdfs.BucketWriter: Renaming hdfs://master:9000/tmp/flume/FlumeData.1652019471500.tmp to hdfs://master:9000/tmp/flume/FlumeData.1652019471500
22/05/08 22:17:57 INFO hdfs.BucketWriter: Creating hdfs://master:9000/tmp/flume/FlumeData.1652019471501.tmp
22/05/08 22:17:57 INFO hdfs.BucketWriter: Closing hdfs://master:9000/tmp/flume/FlumeData.1652019471501.tmp
22/05/08 22:17:57 INFO hdfs.BucketWriter: Renaming hdfs://master:9000/tmp/flume/FlumeData.1652019471501.tmp to hdfs://master:9000/tmp/flume/FlumeData.1652019471501
22/05/08 22:17:57 INFO hdfs.BucketWriter: Creating hdfs://master:9000/tmp/flume/FlumeData.1652019471502.tmp
22/05/08 22:17:57 INFO hdfs.BucketWriter: Closing hdfs://master:9000/tmp/flume/FlumeData.1652019471502.tmp
22/05/08 22:17:57 INFO hdfs.BucketWriter: Renaming hdfs://master:9000/tmp/flume/FlumeData.1652019471502.tmp to hdfs://master:9000/tmp/flume/FlumeData.1652019471502
22/05/08 22:17:57 INFO hdfs.BucketWriter: Creating hdfs://master:9000/tmp/flume/FlumeData.1652019471503.tmp
22/05/08 22:17:57 INFO hdfs.BucketWriter: Closing hdfs://master:9000/tmp/flume/FlumeData.1652019471503.tmp
22/05/08 22:17:57 INFO hdfs.BucketWriter: Renaming hdfs://master:9000/tmp/flume/FlumeData.1652019471503.tmp to hdfs://master:9000/tmp/flume/FlumeData.1652019471503
# 查看 Flume 传输到 HDFS 的文件
[hadoop@master flume]$ hdfs dfs -ls /tmp/flume
-rw-r--r-- 2 hadoop supergroup 1329 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471536
-rw-r--r-- 2 hadoop supergroup 1479 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471537
-rw-r--r-- 2 hadoop supergroup 1360 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471538
-rw-r--r-- 2 hadoop supergroup 1249 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471539
-rw-r--r-- 2 hadoop supergroup 1349 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471540
-rw-r--r-- 2 hadoop supergroup 1550 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471541
-rw-r--r-- 2 hadoop supergroup 1241 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471542
-rw-r--r-- 2 hadoop supergroup 1372 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471543
-rw-r--r-- 2 hadoop supergroup 1362 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471544
-rw-r--r-- 2 hadoop supergroup 1485 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471545
-rw-r--r-- 2 hadoop supergroup 17253 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471546
-rw-r--r-- 2 hadoop supergroup 1296 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471547
-rw-r--r-- 2 hadoop supergroup 1285 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471548
-rw-r--r-- 2 hadoop supergroup 1447 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471549
-rw-r--r-- 2 hadoop supergroup 1363 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471550
-rw-r--r-- 2 hadoop supergroup 1246 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471551
-rw-r--r-- 2 hadoop supergroup 1366 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471552
-rw-r--r-- 2 hadoop supergroup 1630 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471553
-rw-r--r-- 2 hadoop supergroup 1250 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471554
-rw-r--r-- 2 hadoop supergroup 1425 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471555
-rw-r--r-- 2 hadoop supergroup 1329 2022-05-08 22:17 /tmp/flume/FlumeData.1652019471556
……
# 若能查看到 HDFS 上/tmp/flume 目录有传输的数据文件，则表示数据传输成功。
使用 flume-ng agent 命令加载 simple-hdfs-flume.conf 配置信息
查看 Flume 传输到 HDFS 的文件：
浏览器查看：http://master:50070
查看文件：
声明：未经许可，不得转载
【Hadoop】10、Flume组件的更多相关文章
- Flume 组件安装配置
下载和解压 Flume 实验环境可能需要回至第四,五,六章(hadoop和hive),否则后面传输数据可能报错(猜测)! 可 以 从 官 网 下 载 Flume 组 件 安 装 包 , 下 载 地 址 ...
- Hadoop生态圈-Flume的组件之自定义Sink
Hadoop生态圈-Flume的组件之自定义Sink 作者:尹正杰 版权声明:原创作品,谢绝转载!否则将追究法律责任. 本篇博客主要介绍sink相关的API使用两个小案例,想要了解更多关于API的小技 ...
- Hadoop生态圈-Flume的组件之自定义拦截器（interceptor）
Hadoop生态圈-Flume的组件之自定义拦截器(interceptor) 作者:尹正杰 版权声明:原创作品,谢绝转载!否则将追究法律责任. 本篇博客只是举例了一个自定义拦截器的方法,测试字节传输速 ...
- Hadoop生态圈-Flume的组件之sink处理器
Hadoop生态圈-Flume的组件之sink处理器 作者:尹正杰 版权声明:原创作品,谢绝转载!否则将追究法律责任. 一. 二.
- Hadoop生态圈-Flume的组件之拦截器与选择器
Hadoop生态圈-Flume的组件之拦截器与选择器 作者:尹正杰 版权声明:原创作品,谢绝转载!否则将追究法律责任. 本篇博客只是配置的是Flume主流的Interceptors,想要了解更详细 ...
- Hadoop生态圈-Flume的主流source源配置
Hadoop生态圈-Flume的主流source源配置 作者:尹正杰 版权声明:原创作品,谢绝转载!否则将追究法律责任. 本篇博客只是配置的是Flume主流的Source,想要了解更详细的配置信息请参 ...
- Hadoop生态圈-flume日志收集工具完全分布式部署
Hadoop生态圈-flume日志收集工具完全分布式部署 作者:尹正杰 版权声明:原创作品,谢绝转载!否则将追究法律责任. 目前为止,Hadoop的一个主流应用就是对于大规模web日志的分析和处理 ...
- hadoop伪分布式组件安装
一.版本建议 Centos V7.5 Java V1.8 Hadoop V2.7.6 Hive V2.3.3 Mysql V5.7 Spark V2.3 Scala V2.12.6 Flume V1. ...
- Hadoop生态圈-Flume的主流Sinks源配置
Hadoop生态圈-Flume的主流Sinks源配置 作者:尹正杰 版权声明:原创作品,谢绝转载!否则将追究法律责任. 本篇博客只是配置的是Flume主流的Sinks,想要了解更详细的配置信息请参考官 ...
随机推荐
- Springboot 抛出Failed to determine a suitable driver class异常原因
SpringBoot项目,已经依赖了MySQL驱动,却还是无法启动,通过问题排除,如果是启动项目,那么pom值 <packaging>pom</packaging> Faile ...
- python中一切皆对象的理解
一切皆对象? 学过Java都知道,Java中有8个基本类型,其余的都是继承自Object类的引用类型.方法内的基本类型变量会在栈上分配,而引用类型会通过逃逸分析来决定在栈上分配或堆上分配.Java中关 ...
- c++中“::”和“:”啥意思
c++中"::"和":"啥意思 (1)"::" 1)类作用域操作符."::"指明了成员函数所属的类.如:M::f(s) ...
- css边距重叠的解决方案
** css防止边距重叠的方法 ** 今天整理了一下用css防止边距重叠的几种方法先假设一组dom结构 <div class="parent"> <div cla ...
- 【uniapp 开发】手绘签名组件
代码如下: <template> <view class="content"> <button type="primary" @t ...
- 【Android开发】URL[] 转成 bitmap[]
public static Bitmap[] getBitmapFromURL(String[] path) throws MalformedURLException { Bitmap[] b = n ...
- jsp笔记---标签
<meta>标签 <meta> 标签提供了 HTML 文档的元数据.元数据不会显示在客户端,但是会被浏览器解析. META元素通常用于指定网页的描述,关键词,文件的最后修改时间 ...
- Oracle中between 和 in
select * from test_s where id between 2 and 12; between 就是左右全闭区间. SELECT columnsFROM tablesWHERE col ...
- script标签中defer和async的区别（稀土掘金学习）
如果没有defer或async属性,浏览器会立即加载并执行相应的脚本.它不会等待后续加载的文档元素,读取到就会开始加载和执行,这样就阻塞了后续文档的加载. 下图可以直观的看出三者之间的区别: 其中蓝色 ...
- git概述
学习资料来源-人家写得比我好 #视频教程: https://www.bilibili.com/video/BV1vy4y1s7k6?spm_id_from=pageDriver #文档教程 https ...