随着互联网技术的发展,原来的单机发展到多机再到大规模集群,nginx,tomcat,openStack,docker容器等等,一个系统由大量的服务构成,其中每个应用与服务的日志分析管理也变得越来越重要。本文将介绍如何使用fd+es+ka搭建日志收集系统。
关于具体介绍请参考官网: Fluentd: http://www.fluentd.org/ Elasticsearch: https://www.elastic.co/products/elasticsearch Kibana:https://www.elastic.co/products/kibana
对于为什么没有采用ELK的搭配不在本文讨论范围,有兴趣的可以自己了解下。
Fluentd的官方文档介绍的非常详细,如果在今后的使用过程当中遇到问题可直接查阅官方文档http://docs.fluentd.org/。
step1 安装
$ curl -L https://toolbelt.treasuredata.com/sh/install-redhat-td-agent2.sh | sh 如果无法使用在线安装可手动下载rpm包进行安装。本文使用版本td-agent-2.3.0-0.el6.x86_64.rpmstep2 启动
$ /etc/init.d/td-agent start Starting td-agent: [ OK ] $ /etc/init.d/td-agent status td-agent (pid 21678) is running...关于更详细的安装请参考 http://docs.fluentd.org/articles/install-by-rpm#
启动后可看到类似如下数据
[INFO ][node ] [Alexander Lexington] version[1.7.1], pid[1078], build[b88f43f/2015-07-29T09:54:16Z] [INFO ][node ] [Alexander Lexington] initializing ... [INFO ][plugins ] [Alexander Lexington] loaded [], sites [] [INFO ][env ] [Alexander Lexington] using [1] data paths, mounts [[/ (/dev/mapper/VolGroup-lv_root)]], net usable_space [28.8gb], net total_space [35gb], types [ext4] [INFO ][node ] [Alexander Lexington] initialized [INFO ][node ] [Alexander Lexington] starting ... [INFO ][transport ] [Alexander Lexington] bound_address {inet[/0:0:0:0:0:0:0:0:9300]}, publish_address {inet[/10.22.205.101:9300]} [INFO ][discovery ] [Alexander Lexington] elasticsearch/6usA_fpmSeiW5GpYOZgIRQ [INFO ][cluster.service ] [Alexander Lexington] new_master [Alexander Lexington][6usA_fpmSeiW5GpYOZgIRQ][rabbitmq-101][inet[/10.22.205.101:9300]], reason: zen-disco-join (elected_as_master) [INFO ][http ] [Alexander Lexington] bound_address {inet[/0:0:0:0:0:0:0:0:9200]}, publish_address {inet[/10.22.205.101:9200]} [INFO ][node ] [Alexander Lexington] started [INFO ][gateway ] [Alexander Lexington] recovered [0] indices into cluster_state插件为可选安装,如果只是作为一般的用户可不必要安装,如果需要对日志数据进行挖掘与分析及es使用集群相关功能,则建议安装。
(1)kopf插件 集群资源查询&数据查询
./elasticsearch-1.7.1/bin/plugin install lmenezes/elasticsearch-kopf/1.0 如果无法在线安装,可手动下载zip文件并解压到以下路径即可 ./elasticsearch-1.7.1/plugins/kopf 安装完成后访问:http://localhost:9200/_plugin/kopfgit地址:https://github.com/lmenezes/elasticsearch-kopf 注意:kopf与es是有对应版本的安装时请安装对应版本,本文为1X。具体kopf的使用不在本文介绍之列,感兴趣的同学请google
(2)head插件主要用于es的数据查询一般配合kopf使用
./elasticsearch-1.7.1/bin/plugin install mobz/elasticsearch-head 如果无法在线安装,可手动下载zip文件并解压到以下路径即可 ./elasticsearch-1.7.1/plugins/head 安装完成后访问:http://localhost:9200/_plugin/head注意:kibana只是一个前端展示平台需要es作为数据源,所以在启动ka的之前请确认es的连接地址是否正确!
./kibana-4.1.4-linux-x64/config/kibana.yml # The host to bind the server to. host: "0.0.0.0" # The Elasticsearch instance to use for all your queries. elasticsearch_url: "http://localhost:9200"修改elasticsearch_url,其他参数修改请具体参看yml的配置文件。
OK到这里三大软件已经安装完成了是否可以进行日志的采集了?No,因为需要对es支持,所以接下来我们需要安装fluentd的es插件。
(1)安装fluent-plugin-elasticsearch
/usr/sbin/td-agent-gem install fluent-plugin-elasticsearch (2)安装fluentd type 插件 /usr/sbin/td-agent-gem install fluent-plugin-typecast (2)安装secure-forward 插件(非必须但常用) /usr/sbin/td-agent-gem install fluent-plugin-secure-forward 注意:对于插件的安装大家可能会遇到问题,就是gem源始终连接不上。。。没办法在天朝。这里可以使用taobao的源代替官方的源。 sudo gem sources -l sudo gem sources -r http://rubygems.org sudo gem sources -r https://rubygems.org sudo gem sources -a https://ruby.taobao.org/如果gem都没有安装怎么办?那就安装喽~
wget http://production.cf.rubygems.org/rubygems/rubygems-2.2.2.tgz ... ruby setup.rb到这里可算把软件和相关插件安装完毕了,下面将以采集nginx access日志为例。
设备清单
虚机101centos6.5: nginx所在服务器,es,fd,ka所在服务器 client agent虚机102centos6.5 :fd及插件所在服务器server agent,负责文件存储nginx日志,并转发101es存储(1)修改虚机101 fluentd 配置
/etc/td-agent/td-agent.conf 添加tail 源 <source> type tail path /var/log/nginx/access.log format /^(?<remote>[^ ]*) - - \[(?<time>[^\]]*)\] "(?<method>\S+)(?: +(?<path>[^\"]*) +\S*)?" (?<status>[^ ]*) (?<body_bytes_sent>[^ ]*) "(?<http_referer>[^\"]*)" ClientVersion "(?<clientVersion>[^ ]*)" "(?<userAgent>[^\"]*)" "(?<remoteHost>[^ ]*)" "(?<http_x_forwarded_for>[^\"]*)" upstream_response_time "(?<upstream_response_time>[^ ]*)" request_time "(?<request_time>[^ ]*)"\s$/ time_format %d/%b/%Y:%H:%M:%S %z types remote:ip,time:time,method:string,path:string,status:integer:body_bytes_sent:integer,http_referer:string,userAgent:string,remoteHost:string,http_x_forwarded_for:string,upstream_response_time:string,request_time:float tag 101nginx.access.log pos_file /var/log/td-agent/pos/nginx.access.log.pos </source> 添加tag match 把采集的日志转发到虚机102 <match *.access.log> type forward flush_interval 60s buffer_type file buffer_path /var/log/td-agent/buffer/* <server> host 10.22.205.102 port 24224 </server> </match>注意:相关配置参数具体参考fluentd官方文档,再次赞下fluentd文档写的很详细很好!关于日志format的正则可以使用http://fluentular.herokuapp.com/ 来测试格式化的正确性
(2)修改虚机102 fluentd 配置
/etc/td-agent/td-agent.conf 1 102只做转发存储,顾只配置match <match *.access.log> type copy <store> type file path /var/log/swq_test/nginx-access/ time_slice_format ./nginx-access/%Y/%m/%d/%Y%m%d%H.nginx.access compress gzip flush_interval 10m time_format %Y-%m-%dT%H:%M:%S%z buffer_path /var/log/swq_test/buffer/nginx_access_buffer buffer_type file buffer_chunk_limit 50m </store> <store> type elasticsearch host 10.22.205.101 port 9200 include_tag_key true tag_key @log_name logstash_format true flush_interval 10s </store> </match>这里102对匹配到的日志做了2个操作:(1). 以时间文件夹路径本地存储nginx日志 (2). 转发获得的日志到101的es上。
重启101,102 fluentd 大功告成!
/etc/init.d/td-agent restartkibana discover:
kibana demo dashboard:
