章节内容
上一节完成了如下的内容:
- 编写Agent Conf配置文件
- 收集Hive数据
- 汇聚到HDFS中
- 测试效果
背景介绍
这里是三台公网云服务器,每台 2C4G,搭建一个Hadoop的学习环境,供我学习。
之前已经在 VM 虚拟机上搭建过一次,但是没留下笔记,这次趁着前几天薅羊毛的3台机器,赶紧尝试在公网上搭建体验一下。
- 2C4G 编号 h121
- 2C4G 编号 h122
- 2C2G 编号 h123
文档推荐
除了官方文档以外,这里有一个写的很好的中文文档:
https://flume.liyifeng.org/
监控目录
业务需求
- 想要监控指定目录 收集信息并上传到HDFS中
Source
选择 spooldir,因为 spooldir 能够保证数据不丢失,且能够进行断点续传,但是延迟较高,不能实时监控。
Channel
选择 memory
Sink
选择 HDFS
需要注意
- 拷贝到 spool 目录下的文件 不可以再打开编辑
- 无法监控子目录的文件夹变动
- 被监控文件夹每500毫秒 扫描一次文件变动
- 适合用于同步新文件,但不适合对实时追加日志的文件进行监听并同步
配置文件
shell
cd /opt/wzk/flume_test
vim flume_spooldir-hdfs.conf
我们需要写入如下内容
shell
# Name the components on this agent
a3.sources = r3
a3.channels = c3
a3.sinks = k3
# Describe/configure the source
a3.sources.r3.type = spooldir
# 注意这里的文件夹 换成自己的!!!
a3.sources.r3.spoolDir = /opt/wzk/upload
a3.sources.r3.fileSuffix = .COMPLETED
a3.sources.r3.fileHeader = true
# 忽略以.tmp结尾的文件,不上传
a3.sources.r3.ignorePattern = ([^ ]*\.tmp)
# Use a channel which buffers events in memory
a3.channels.c3.type = memory
a3.channels.c3.capacity = 10000
a3.channels.c3.transactionCapacity = 500
# Describe the sink
a3.sinks.k3.type = hdfs
# 注意修改成你自己的IP!!!
a3.sinks.k3.hdfs.path = hdfs://h121.wzk.icu:9000/flume/upload/%Y%m%d/%H%M
# 上传文件的前缀
a3.sinks.k3.hdfs.filePrefix = upload-
# 是否使用本地时间戳
a3.sinks.k3.hdfs.useLocalTimeStamp = true
# 积攒500个Event,flush到HDFS一次
a3.sinks.k3.hdfs.batchSize = 500
# 设置文件类型
a3.sinks.k3.hdfs.fileType = DataStream
# 60秒滚动一次
a3.sinks.k3.hdfs.rollInterval = 60
# 128M滚动一次
a3.sinks.k3.hdfs.rollSize = 134217700
# 文件滚动与event数量无关
a3.sinks.k3.hdfs.rollCount = 0
# 最小冗余数
a3.sinks.k3.hdfs.minBlockReplicas = 1
# Bind the source and sink to the channel
a3.sources.r3.channels = c3
a3.sinks.k3.channel = c3
启动Agent
shell
$FLUME_HOME/bin/flume-ng agent --name a3 \
--conf-file flume-spooldir-hdfs.conf \
-Dflume.root.logger=INFO,console
测试效果
Flume
shell
cd /opt/wzk/upload
vim 1.txt
随便向其中写入一些内容,并保存,可以看到Flume已经有反应了。
HDFS
查看HDFS,也已经有内容了
采集双写
这里业务上需要:
- Flume将数据写入本地
- Flume将数据写入HDFS
分析实现
- 需要多个Agent级联实现
- Source选择taildir
- Channel选择memory
- 最终的Sink分别选择HDFS,file_roll
配置文件1
配置文件包含如下内容:
- 1个 taildir source
- 2个 memory channel
- 2个 avro sink
新建文件
shell
vim flume-taildir-avro.conf
写入如下内容
shell
# Name the components on this agent
a1.sources = r1
a1.sinks = k1 k2
a1.channels = c1 c2
# 将数据流复制给所有channel
a1.sources.r1.selector.type = replicating
# source
a1.sources.r1.type = taildir
# 记录每个文件最新消费位置
a1.sources.r1.positionFile = /root/flume/taildir_position.json
a1.sources.r1.filegroups = f1
# 备注:.*log 是正则表达式;这里写成 *.log 是错误的
a1.sources.r1.filegroups.f1 = /tmp/root/.*log
# sink
a1.sinks.k1.type = avro
a1.sinks.k1.hostname = linux123
a1.sinks.k1.port = 9091
a1.sinks.k2.type = avro
a1.sinks.k2.hostname = linux123
a1.sinks.k2.port = 9092
# channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 10000
a1.channels.c1.transactionCapacity = 500
a1.channels.c2.type = memory
a1.channels.c2.capacity = 10000
a1.channels.c2.transactionCapacity = 500
# Bind the source and sink to the channel
a1.sources.r1.channels = c1 c2
a1.sinks.k1.channel = c1
a1.sinks.k2.channel = c2
配置文件2
配置文件包含如下内容:
- 1个 avro source
- 1个 memory channel
- 1个 hdfs sink
新建配置文件
shell
vim flume-avro-hdfs.conf
写入如下的内容:
shell
# Name the components on this agent
a2.sources = r1
a2.sinks = k1
a2.channels = c1
# Describe/configure the source
a2.sources.r1.type = avro
a2.sources.r1.bind = linux123
a2.sources.r1.port = 9091
# Describe the channel
a2.channels.c1.type = memory
a2.channels.c1.capacity = 10000
a2.channels.c1.transactionCapacity = 500
# Describe the sink
a2.sinks.k1.type = hdfs
a2.sinks.k1.hdfs.path = hdfs://linux121:8020/flume2/%Y%m%d/%H
# 上传文件的前缀
a2.sinks.k1.hdfs.filePrefix = flume2-
# 是否使用本地时间戳
a2.sinks.k1.hdfs.useLocalTimeStamp = true
# 500个Event才flush到HDFS一次
a2.sinks.k1.hdfs.batchSize = 500
# 设置文件类型,可支持压缩
a2.sinks.k1.hdfs.fileType = DataStream
# 60秒生成一个新的文件
a2.sinks.k1.hdfs.rollInterval = 60
a2.sinks.k1.hdfs.rollSize = 0
a2.sinks.k1.hdfs.rollCount = 0
a2.sinks.k1.hdfs.minBlockReplicas = 1
# Bind the source and sink to the channel
a2.sources.r1.channels = c1
a2.sinks.k1.channel = c1
配置文件3
配置文件包含如下内容:
- 1个 avro source
- 1个 memory channel
- 1个 file_roll sink
新建配置文件
shell
vim flume-avro-file.conf
写入如下的内容
shell
# Name the components on this agent
a3.sources = r1
a3.sinks = k1
a3.channels = c2
# Describe/configure the source
a3.sources.r1.type = avro
a3.sources.r1.bind = linux123
a3.sources.r1.port = 9092
# Describe the sink
a3.sinks.k1.type = file_roll
# 目录需要提前创建好
a3.sinks.k1.sink.directory = /root/flume/output
# Describe the channel
a3.channels.c2.type = memory
a3.channels.c2.capacity = 10000
a3.channels.c2.transactionCapacity = 500
# Bind the source and sink to the channel
a3.sources.r1.channels = c2
a3.sinks.k1.channel = c2
启动Agent1
shell
$FLUME_HOME/bin/flume-ng agent --name a3 \
--conf-file ~/conf/flume-avro-file.conf \
-Dflume.root.logger=INFO,console &
启动Agent2
shell
$FLUME_HOME/bin/flume-ng agent --name a2 \
--conf-file ~/conf/flume-avro-hdfs.conf \
-Dflume.root.logger=INFO,console &
启动Agent3
shell
$FLUME_HOME/bin/flume-ng agent --name a1 \
--conf-file ~/conf/flume-taildir-avro.conf \
-Dflume.root.logger=INFO,console &
Hive测试
shell
hive -e "show databases;"