#################
# netcat 配置 #
#################
#解压、重命名
cd /opt/download
tar -zxvf netcat-0.7.1.tar.gz -C /opt/software
#可能报错需要将gcc升至5.3以上
yum -y install centos-release-scl
yum -y install devtoolset-9-gcc devtoolset-9-gcc-c++ devtoolset-9-binutils
scl enable devtoolset-9 bash #临时设置
echo "source /opt/rh/devtoolset-9/enable" >> /etc/profile #永久设置
gcc -v
#--------------------------------------------------
gcc version 9.3.1 20200408 (Red Hat 9.3.1-2) (GCC)
#--------------------------------------------------
cd /opt/software/netcat-0.7.1/
./configure --prefix=/usr/local/netcat
make && make install
cd /usr/local/netcat
ls
#------------------
bin info man share
#------------------
#配置环境变量
vim /etc/profile.d/my.sh
#---------------------------------------------------------
#netcat071
export NETCAT_HOME=/usr/local/netcat
export PATH=$PATH:$NETCAT_HOME/bin
#-----------------------------------------------------------
source /etc/profile
nc localhost 9999
################
# flume 安装 #
################
#解压、重命名
cd /opt/download
tar -zxvf apache-flume-1.9.0-bin.tar.gz -C /opt/software
mv apache-flume-1.9.0 flume190
#配置环境变量
vim /etc/profile.d/my.sh
#----------------------------------------
#flume190
export FLUME_HOME=/opt/software/flume190
export PATH=$PATH:$FLUME_HOME/bin
#----------------------------------------
source /etc/profile
#修改jdk的路径
cd /opt/software/flume190/conf/
mv flume-env.sh.template flume-env.sh
vim flume-env.sh
#-----------------------------------
#修改
export JAVA_HOME=/opt/software/jdk8
#-----------------------------------
# flume-ng 运行异常时解决方法如下
#1.hdfs-sink时出错
#更新guava包和hadoop一致
cd /opt/software/flume-1.9.0
ls lib|grep guava
#-------------------
guava-11.0.2.jar
#-------------------
rm -f lib/guava-11.0.0.jar
find /opt/software/hadoop313/share/hadoop/hdfs/lib -name guava*
#------------------------------------------------------------------
/opt/software/hadoop313/share/hadoop/hdfs/lib/guava-27.0-jre.jar
#------------------------------------------------------------------
cp /opt/software/hadoop313/share/hadoop/hdfs/lib/guava-27.0-jre.jar lib/
#拷贝hadoop架包到flume里
cp /opt/software/hadoop313/share/hadoop/client/*.jar /opt/software/flume-1.9.0/lib/
cp /opt/software/hadoop313/share/hadoop/common/*.jar /opt/software/flume-1.9.0/lib/
cp /opt/software/hadoop313/share/hadoop/hdfs/*.jar /opt/software/flume-1.9.0/lib/
cp /opt/software/hadoop313/share/hadoop/mapreduce/*.jar /opt/software/flume-1.9.0/lib/
cp /opt/software/hadoop313/share/hadoop/tools/*/*.jar /opt/software/flume-1.9.0/lib/
cp /opt/software/hadoop313/share/hadoop/yarn/*.jar /opt/software/flume-1.9.0/lib/
cp /opt/software/hadoop313/share/hadoop/common/*/*.jar /opt/software/flume-1.9.0/lib/
#2.报错:(java.lang.OutOfMemoryError: GC overhead limit exceeded)
vim /opt/software/flume190/bin/flume-ng
#----------------------
#修改
JAVA_OPTS="-Xmx2048m"
#----------------------
#官网
https://flume.apache.org/releases/content/1.9.0/FlumeUserGuide.html
# flume可以设置文件:
#rollSize
默认值:1024,当临时文件达到该大小(单位:bytes)时,滚动成目标文件。
如果设置成0,则表示不根据临时文件大小来滚动文件。
#rollCount
默认值:10,当events数据达到该数量时候,将临时文件滚动成目标文件,
如果设置成0,则表示不根据events数据来滚动文件。
#round
默认值:false,是否启用时间上的"舍弃",类似于四舍五入,
如果启用,则会影响除了%t的其他所有时间表达式。
#roundValue
默认值:1,时间上进行"舍弃"的值。
#roundUnit
默认值:seconds,时间上进行"舍弃"的单位,包含:second,minute,hour
当设置了round、roundValue、roundUnit参数时,需要在sink指定的HDFS路径上指定按照时间生成的目录的格式,
如有需求,每采集1小时就在HDFS目录上生成一个目录,里面存放这1小时内采集到的数据。
=========================================================================================================================
# 1、模拟 netcat写入数据,flume接收
cd ~
mkdir flume
vim flume_avro_memory_log.log
{
#common
a.sources = s1
a.channels = c1
a.sinks = k1
#source
a.sources.s1.type = netcat
a.sources.s1.bind = localhost
a.sources.s1.port = 9999
#channel
a.channels.c1.type = memory
a.channels.c1.capacity = 1024
a.channels.c1.transactionCapacity = 128
#sink
a.sinks.k1.type = logger
#join
a.sources.s1.channels = c1
a.sinks.k1.channel = c1
}
flume-ng agent --name a --conf /opt/software/flume190/conf/ --conf-file ~/flume/flume_avro_memory_log.log -Dflume.root.logger=INFO,console
#或者
flume-ng agent -n a -c /opt/software/flume190/conf/ -f ~/flume/flume_avro_memory_log.log -Dflume.root.logger=INFO,console
# 2、模拟 文件
cd flume/
mkdir spooldir #spooldir文件夹下拖入文件
cd flume/
vim flume_spooldir_memory_log.log
{
#common
a.sources = s1
a.channels = c1
a.sinks = k1
#source
a.sources.s1.type = spooldir
a.sources.s1.spoolDir = /root/flume/spooldir
a.sources.s1.fileHeader = true
#channel
a.channels.c1.type = memory
a.channels.c1.capacity = 1024
a.channels.c1.transactionCapacity = 128
#sink
a.sinks.k1.type = logger
#join
a.sources.s1.channels = c1
a.sinks.k1.channel = c1
}
flume-ng agent -n a -c /opt/software/flume190/conf/ -f ~/flume/flume_spooldir_memory_log.log -Dflume.root.logger=INFO,console
# 3、模拟 文件写入hdfs
cd flume/
mkdir spooldir #spooldir文件夹下拖入文件
cd flume/
vim flume_spooldir_memory_hdfs.log
{
#common
a.sources = s1
a.channels = c1
a.sinks = k1
#source
a.sources.s1.type = spooldir
a.sources.s1.spoolDir = /root/flume/spooldir
a.sources.s1.fileHeader = true
#channel
a.channels.c1.type = memory
a.channels.c1.capacity = 201326592
a.channels.c1.transactionCapacity = 134217728
#sink
a.sinks.k1.type = hdfs
a.sinks.k1.hdfs.path = hdfs://single1:9000/flume/dir_202228_01
a.sinks.k1.hdfs.filePrefix = tags-events-
a.sinks.k1.hdfs.round = true
a.sinks.k1.hdfs.roundValue = 10
a.sinks.k1.hdfs.roundUnit = second
#join
a.sources.s1.channels = c1
a.sinks.k1.channel = c1
}
flume-ng agent -n a -c /opt/software/flume190/conf/ -f ~/flume/flume_spooldir_memory_hdfs.log -Dflume.root.logger=INFO,console
# 4、模拟 写入hdfs
cd flume/
mkdir spooldir #spooldir文件夹下拖入文件
cd flume/
vim flume_taildir_memory_hdfs.log
{
#common
a.sources=s1
a.channels=c1
a.sinks=k1
#source
a.sources.s1.type=TAILDIR
a.sources.s1.positionFile=/root/flume/position_file/pf_01.json
a.sources.s1.filegroups=f1
a.sources.s1.filegroups.f1=/root/flume/spooldir/.*.csv
a.sources.s1.fileHeader=true
a.sources.s1.maxBatchCount=1000
#channel
a.channels.c1.type=memory
a.channels.c1.capacity=16777216
a.channels.c1.transactionCapacity=8388608
#sink
a.sinks.k1.type = hdfs
a.sinks.k1.hdfs.path = hdfs://single1:9000/flume/dir_202228_02
a.sinks.k1.hdfs.filePrefix = tags-events-
a.sinks.k1.hdfs.round = true
a.sinks.k1.hdfs.roundValue = 10
a.sinks.k1.hdfs.roundUnit = second
a.sinks.k1.hdfs.rollInterval = 0
a.sinks.k1.hdfs.rollCount = 0
a.sinks.k1.hdfs.rollSize = 33554432
#join
a.sources.s1.channels = c1
a.sinks.k1.channel = c1
}
flume-ng agent -n a -c /opt/software/flume190/conf/ -f ~/flume/flume_taildir_memory_hdfs.log -Dflume.root.logger=INFO,console
# 5、模拟多级联动
#single1执行
vim flume_link_netcat_memory_avro.log
{
#common
a.sources=s1
a.channels=c1
a.sinks=k1
#source
a.sources.s1.type=netcat
a.sources.s1.bind=single1
a.sources.s1.port=9999
#channel
a.channels.c1.type=memory
a.channels.c1.capacity=256
a.channels.c1.transactionCapacity=128
#sink
a.sinks.k1.type=avro
a.sinks.k1.hostname=single01
a.sinks.k1.port=44444
#join
a.sources.s1.channels = c1
a.sinks.k1.channel = c1
}
flume-ng agent -n a -c /opt/software/flume190/conf/ -f ~/flume/flume_link_netcat_memory_avro.log -Dflume.root.logger=INFO,console
#single01执行
vim flume_link_avro_memory_log.log
{
#common
a.sources=s1
a.channels=c1
a.sinks=k1
#source
a.sources.s1.type=avro
a.sources.s1.bind=single01
a.sources.s1.port=44444
#channel
a.channels.c1.type=memory
a.channels.c1.capacity=256
a.channels.c1.transactionCapacity=128
#sink
a.sinks.k1.type=logger
#join
a.sources.s1.channels = c1
a.sinks.k1.channel = c1
}
flume-ng agent -n a -c /opt/software/flume190/conf/ -f ~/flume/flume_link_avro_memory_log.log -Dflume.root.logger=INFO,console
# 6、模拟扇入(single1作为接收,single1和single01同时作为数据发送源)
#服务端-接收single1
vim flume_fanin1_avro_memory_log.log #single1执行 入口先启动
{
#common
a.sources=s1 s2
a.channels=c1
a.sinks=k1
#source
a.sources.s1.type=avro
a.sources.s1.bind=single1
a.sources.s1.port=44444
#channel
a.channels.c1.type=memory
a.channels.c1.capacity=256
a.channels.c1.transactionCapacity=128
#sink
a.sinks.k1.type=logger
#join
a.sources.s1.channels = c1
a.sinks.k1.channel = c1
}
flume-ng agent -n a -c /opt/software/flume190/conf/ -f ~/flume/flume_fanin1_avro_memory_log.log -Dflume.root.logger=INFO,console
#链接发送-客户端single1
vim flume_fanin1_netcat_memory_avro.log
{
#common
a.sources=s1
a.channels=c1
a.sinks=k1
#source
a.sources.s1.type=netcat
a.sources.s1.bind=single1
a.sources.s1.port=9999
#channel
a.channels.c1.type=memory
a.channels.c1.capacity=256
a.channels.c1.transactionCapacity=128
#sink
a.sinks.k1.type=avro
a.sinks.k1.hostname=single1
a.sinks.k1.port=44444
#join
a.sources.s1.channels = c1
a.sinks.k1.channel = c1
}
flume-ng agent -n a -c /opt/software/flume190/conf/ -f ~/flume/flume_fanin1_netcat_memory_avro.log -Dflume.root.logger=INFO,console
#链接发送-客户端single01
vim flume_fanin01_netcat_memory_avro.log
{
#common
a.sources=s1
a.channels=c1
a.sinks=k1
#source
a.sources.s1.type=netcat
a.sources.s1.bind=single01
a.sources.s1.port=9999
#channel
a.channels.c1.type=memory
a.channels.c1.capacity=256
a.channels.c1.transactionCapacity=128
#sink
a.sinks.k1.type=avro
a.sinks.k1.hostname=single1
a.sinks.k1.port=44444
#join
a.sources.s1.channels = c1
a.sinks.k1.channel = c1
}
flume-ng agent -n a -c /opt/software/flume190/conf/ -f ~/flume/flume_fanin01_netcat_memory_avro.log -Dflume.root.logger=INFO,console
# 7、模拟扇出
#创建kafka主题
kafka-topics.sh --bootstrap-server single1:9092 --create --topic flume_kafka_channel_fanout_20220209_01 --partitions 1 --replication-factor 1
#创建kafka消费者
kafka-console-consumer.sh --bootstrap-server single1:9092 --topic flume_kafka_channel_fanout_20220209_01 --from-beginning
#创建hbase表
create "kb16nb:flume_hbase_sink_20220209","tags"
vim flume_fanout_taildir_file_log_hbase_kafka.log
{
#common
a.sources=s1
a.channels=c1 c2 c3
a.sinks=k1 k2 k3
#source
a.sources.s1.type=TAILDIR
a.sources.s1.positionFile=/root/flume/position_file/pf_01.json
a.sources.s1.filegroups=f1
a.sources.s1.filegroups.f1=/root/flume/spooldir/.*.csv
a.sources.s1.fileHeader=true
a.sources.s1.maxBatchCount=1000
#channel
a.channels.c1.type=memory
a.channels.c1.capacity=8192
a.channels.c1.transactionCapacity=4096
a.channels.c2.type=org.apache.flume.channel.kafka.KafkaChannel
a.channels.c2.kafka.bootstrap.servers=single1:9092
a.channels.c2.kafka.topic=flume_kafka_channel_fanout_20220209_01
a.channels.c2.kafka.consumer.group.id=flume_kafka_channel_fanout_20220209_01
a.channels.c3.type=file
a.channels.c3.checkpointDir=/root/flume/checkpoint
a.channels.c3.dataDirs=/root/flume/datadir
a.channels.c3.capacity=8192
a.channels.c3.transactionCapacity=4096
a.channels.c3.checkpointInterval=10000
#sink
a.sinks.k1.type=logger
a.sinks.k2.type=hbase
a.sinks.k2.table=kb16nb:flume_hbase_sink_20220209
a.sinks.k2.columnFamily=tags
a.sinks.k2.zookeeperQuorum=single1:2181
a.sinks.k2.batchSize=50
a.sinks.k2.serializer.regex=.*?@(.*?),(.*?),(.*?),(.*?)
a.sinks.k2.serializer=org.apache.flume.sink.hbase.RegexHbaseEventSerializer
a.sinks.k2.serializer.colNames=userId,movieId,tag,timestamp
a.sinks.k3.type=org.apache.flume.sink.kafka.KafkaSink
a.sinks.k3.kafka.bootstrap.servers=single1:9092
a.sinks.k3.kafka.topic=flume_kafka_sink_fanout_20220209_01
a.sinks.k3.flumeBatchSize=50
#join
a.sources.s1.channels=c1 c2 c3
a.sinks.k1.channel=c1
a.sinks.k2.channel=c2
a.sinks.k3.channel=c3
}
flume-ng agent -n a -c /opt/software/flume190/conf/ -f ~/flume/flume_fanout_taildir_file_log_hbase_kafka.log -Dflume.root.logger=INFO,console