nginx+flume+hdfs搭建实时日志收集系统

1、配置nginx.conf，添加以下配置

http {
#配置日志格式
log_format lf '$remote_addr^A$msec^A$http_host^A$request_uri';
server {
listen       80;
server_name  localhost;
location / {
access_log /home/bxp/Documents/install/tengine-2.2.0/log/nginx/access.log lf;
root   html;
}
}
}

2、重启nginx

systemctl restart nginx

3、创建flume agent文件，内容如下

#define agent
agent.sources = r2
agent.channels = c2
agent.sinks = k2

# defined source
agent.sources.r2.type = exec
agent.sources.r2.command = tail -f /home/bxp/Documents/install/tengine-2.2.0/log/nginx/access.log
agent.sources.r2.shell = /bin/bash -c

# defined channel
agent.channels.c2.type = memory
#设置channel的容量
agent.channels.c2.capacity = 1000
#设置sink每次从channel中拉取的event的数量
agent.channels.c2.transactionCapacity = 100

# defined sinks
agent.sinks.k2.type = hdfs
agent.sinks.k2.hdfs.path = hdfs://hadoop-series.bxp.com:8020/user/bxp/flume/tracker/%Y/%m/%d
agent.sinks.k2.hdfs.fileType = DataStream
agent.sinks.k2.hdfs.writeFormat = Text
agent.sinks.k2.hdfs.batchSize = 10
agent.sinks.k2.hdfs.useLocalTimeStamp = true

#bind the sources and sink to the channel
agent.sources.r2.channels = c2
agent.sinks.k2.channel = c2

4、将hdfs和flume集成的jar拷贝到flume的lib目录下，需要的jia如下

{HADOOP_HOME}/share/hadoop/common/lib/commons-configuration-1.6.jar
{HADOOP_HOME}/share/hadoop/common/lib/hadoop-auth-2.6.0-cdh5.10.0.jar
{HADOOP_HOME}/share/hadoop/common/hadoop-common-2.6.0-cdh5.10.0.jar
{HADOOP_HOME}/share/hadoop/hdfs/hadoop-hdfs-2.6.0-cdh5.10.0.jar

5、启动HDFS

6、启动flume

bin/flume-ng agent --conf conf/  --name agent --conf-file conf/agent-commad.conf -Dflume.root.logger=DEBUG,console