本文共 6859 字,大约阅读时间需要 22 分钟。
Logstash啊,根据input来监控数据,根据output来使用数据!!!
手把手带你看官方文档(Logstash inputs和Logstash outputs)
关于 Logstash inputs、Logstash outputs和Filter plugins,很多很多,自行去官网看,我这里不多赘述。仅仅拿下面来示范:
Logstash 的input
file input
最常用的input插件是file。
[hadoop@HadoopMaster logstash-2.4.1]$ pwd/home/hadoop/app/logstash-2.4.1[hadoop@HadoopMaster logstash-2.4.1]$ lltotal 164drwxrwxr-x. 2 hadoop hadoop 4096 Mar 27 03:58 bin-rw-rw-r--. 1 hadoop hadoop 102879 Nov 14 10:04 CHANGELOG.md-rw-rw-r--. 1 hadoop hadoop 2249 Nov 14 10:04 CONTRIBUTORS-rw-rw-r--. 1 hadoop hadoop 5084 Nov 14 10:07 Gemfile-rw-rw-r--. 1 hadoop hadoop 23015 Nov 14 10:04 Gemfile.jruby-1.9.lockdrwxrwxr-x. 4 hadoop hadoop 4096 Mar 27 03:58 lib-rw-rw-r--. 1 hadoop hadoop 589 Nov 14 10:04 LICENSE-rw-rw-r--. 1 hadoop hadoop 46 Mar 27 05:30 logstash-simple.conf-rw-rw-r--. 1 hadoop hadoop 149 Nov 14 10:04 NOTICE.TXTdrwxrwxr-x. 4 hadoop hadoop 4096 Mar 27 03:58 vendor[hadoop@HadoopMaster logstash-2.4.1]$ vim file_stdout.conf
这个,是可以自定义的。我这里是
path => "/home/hadoop/app.log" 或者 path => [ "/home/hadoop/app", "/home/hadoop/*.log" ]
input { file { path => "/home/hadoop/app.log" }}filter { }output { stdout {}}
我这里是, 监控/home/hadoop/app.log这个文件的变化。
[hadoop@HadoopMaster logstash-2.4.1]$ pwd/home/hadoop/app/logstash-2.4.1[hadoop@HadoopMaster logstash-2.4.1]$ lltotal 168drwxrwxr-x. 2 hadoop hadoop 4096 Mar 27 03:58 bin-rw-rw-r--. 1 hadoop hadoop 102879 Nov 14 10:04 CHANGELOG.md-rw-rw-r--. 1 hadoop hadoop 2249 Nov 14 10:04 CONTRIBUTORS-rw-rw-r--. 1 hadoop hadoop 107 Mar 27 05:55 file_stdout.conf-rw-rw-r--. 1 hadoop hadoop 5084 Nov 14 10:07 Gemfile-rw-rw-r--. 1 hadoop hadoop 23015 Nov 14 10:04 Gemfile.jruby-1.9.lockdrwxrwxr-x. 4 hadoop hadoop 4096 Mar 27 03:58 lib-rw-rw-r--. 1 hadoop hadoop 589 Nov 14 10:04 LICENSE-rw-rw-r--. 1 hadoop hadoop 46 Mar 27 05:30 logstash-simple.conf-rw-rw-r--. 1 hadoop hadoop 149 Nov 14 10:04 NOTICE.TXTdrwxrwxr-x. 4 hadoop hadoop 4096 Mar 27 03:58 vendor[hadoop@HadoopMaster logstash-2.4.1]$ bin/logstash -f file_stdout.conf Settings: Default pipeline workers: 1Pipeline main started
重新打开,另外一个HadoopMaster界面。
[hadoop@HadoopMaster ~]$ pwd/home/hadoop[hadoop@HadoopMaster ~]$ lltotal 48drwxrwxr-x. 12 hadoop hadoop 4096 Mar 27 03:59 app-rw-rw-r--. 1 hadoop hadoop 18 Mar 26 19:59 app.logdrwxrwxr-x. 7 hadoop hadoop 4096 Mar 25 06:34 datadrwxr-xr-x. 2 hadoop hadoop 4096 Oct 31 17:19 Desktopdrwxr-xr-x. 2 hadoop hadoop 4096 Oct 31 17:19 Documentsdrwxr-xr-x. 2 hadoop hadoop 4096 Oct 31 17:19 Downloadsdrwxr-xr-x. 2 hadoop hadoop 4096 Oct 31 17:19 Musicdrwxr-xr-x. 2 hadoop hadoop 4096 Mar 26 20:35 mybeatdrwxr-xr-x. 2 hadoop hadoop 4096 Oct 31 17:19 Picturesdrwxr-xr-x. 2 hadoop hadoop 4096 Oct 31 17:19 Publicdrwxr-xr-x. 2 hadoop hadoop 4096 Oct 31 17:19 Templatesdrwxr-xr-x. 2 hadoop hadoop 4096 Oct 31 17:19 Videos[hadoop@HadoopMaster ~]$ echo bbbbbbb >> app.log [hadoop@HadoopMaster ~]$
[hadoop@HadoopMaster logstash-2.4.1]$ pwd/home/hadoop/app/logstash-2.4.1[hadoop@HadoopMaster logstash-2.4.1]$ lltotal 168drwxrwxr-x. 2 hadoop hadoop 4096 Mar 27 03:58 bin-rw-rw-r--. 1 hadoop hadoop 102879 Nov 14 10:04 CHANGELOG.md-rw-rw-r--. 1 hadoop hadoop 2249 Nov 14 10:04 CONTRIBUTORS-rw-rw-r--. 1 hadoop hadoop 107 Mar 27 05:55 file_stdout.conf-rw-rw-r--. 1 hadoop hadoop 5084 Nov 14 10:07 Gemfile-rw-rw-r--. 1 hadoop hadoop 23015 Nov 14 10:04 Gemfile.jruby-1.9.lockdrwxrwxr-x. 4 hadoop hadoop 4096 Mar 27 03:58 lib-rw-rw-r--. 1 hadoop hadoop 589 Nov 14 10:04 LICENSE-rw-rw-r--. 1 hadoop hadoop 46 Mar 27 05:30 logstash-simple.conf-rw-rw-r--. 1 hadoop hadoop 149 Nov 14 10:04 NOTICE.TXTdrwxrwxr-x. 4 hadoop hadoop 4096 Mar 27 03:58 vendor[hadoop@HadoopMaster logstash-2.4.1]$ bin/logstash -f file_stdout.conf Settings: Default pipeline workers: 1Pipeline main started2017-03-26T22:24:35.897Z HadoopMaster bbbbbbb
其实,这个文件,.sincedb_8f3299d0a5bdb7df6154f681fc150341也会记录。
注意:
第一次读取新文件,不会有.sincedb等这些,默认根据这个start_position去读,若start_position是end,则读最后。若start_position是begin,则读最开始。
若不是第一次读取文件了,重启Logstash,则会有.sincedb文件了,则就转去根据这个.sincedb文件读了。不管start_position是什么,都不起效了。
start_position:指定从什么位置开始读取文件数据,默认是结束位置,也可以指定为从头开始。
注意:start_position仅在该文件从未被监听过的时候起作用,因为logstash在读取文件的时候会记录一个.sincedb文件来跟踪文件的读取位置,当文件被读取过一次之后,下次就会从.sincedb中记录的位置读取,start_position参数就无效了。文件默认在用户目录下。注意一个坑:ignore_older属性,表示忽略老的数据,值默认为86400,表示忽略24小时以前的数据。如果你新监控一个24小时以上没有被修改过的老文件的话,就算把start_position设置为beginning,也无法获取之前的数据。
redis input
[hadoop@HadoopMaster logstash-2.4.1]$ pwd/home/hadoop/app/logstash-2.4.1[hadoop@HadoopMaster logstash-2.4.1]$ lltotal 168drwxrwxr-x. 2 hadoop hadoop 4096 Mar 27 03:58 bin-rw-rw-r--. 1 hadoop hadoop 102879 Nov 14 10:04 CHANGELOG.md-rw-rw-r--. 1 hadoop hadoop 2249 Nov 14 10:04 CONTRIBUTORS-rw-rw-r--. 1 hadoop hadoop 107 Mar 27 05:55 file_stdout.conf-rw-rw-r--. 1 hadoop hadoop 5084 Nov 14 10:07 Gemfile-rw-rw-r--. 1 hadoop hadoop 23015 Nov 14 10:04 Gemfile.jruby-1.9.lockdrwxrwxr-x. 4 hadoop hadoop 4096 Mar 27 03:58 lib-rw-rw-r--. 1 hadoop hadoop 589 Nov 14 10:04 LICENSE-rw-rw-r--. 1 hadoop hadoop 46 Mar 27 05:30 logstash-simple.conf-rw-rw-r--. 1 hadoop hadoop 149 Nov 14 10:04 NOTICE.TXTdrwxrwxr-x. 4 hadoop hadoop 4096 Mar 27 03:58 vendor[hadoop@HadoopMaster logstash-2.4.1]$ vim redis_stdout.conf
input { redis { host => "192.168.80.12" port => 6379 data_type => "list" key => "filebeat" }}filter { }output { stdout {}}
这里,用到redis,不会用的博友,请移步
这里,等我安装好了之后,再来。
Logstash 的output
redis output
[hadoop@HadoopMaster logstash-2.4.1]$ pwd/home/hadoop/app/logstash-2.4.1[hadoop@HadoopMaster logstash-2.4.1]$ lltotal 172drwxrwxr-x. 2 hadoop hadoop 4096 Mar 27 03:58 bin-rw-rw-r--. 1 hadoop hadoop 102879 Nov 14 10:04 CHANGELOG.md-rw-rw-r--. 1 hadoop hadoop 2249 Nov 14 10:04 CONTRIBUTORS-rw-rw-r--. 1 hadoop hadoop 107 Mar 27 05:55 file_stdout.conf-rw-rw-r--. 1 hadoop hadoop 5084 Nov 14 10:07 Gemfile-rw-rw-r--. 1 hadoop hadoop 23015 Nov 14 10:04 Gemfile.jruby-1.9.lockdrwxrwxr-x. 4 hadoop hadoop 4096 Mar 27 03:58 lib-rw-rw-r--. 1 hadoop hadoop 589 Nov 14 10:04 LICENSE-rw-rw-r--. 1 hadoop hadoop 46 Mar 27 05:30 logstash-simple.conf-rw-rw-r--. 1 hadoop hadoop 149 Nov 14 10:04 NOTICE.TXT-rw-rw-r--. 1 hadoop hadoop 155 Mar 27 06:43 redis_stdout.confdrwxrwxr-x. 4 hadoop hadoop 4096 Mar 27 03:58 vendor[hadoop@HadoopMaster logstash-2.4.1]$ vim stdin_es.conf
elasticsearch output
即把Logstash里的数据,写到elasticsearch 集群(这台192.168.80.10里)
hosts=>"192.168.80.00"或 hosts=>["192.168.80.10:9200","192.168.80.11:9200","192.168.80.12:9200"]
1.x中属性名称叫host
默认向es中创建的索引库是logstash-%{+YYYY.MM.dd},可以利用es中的索引模板特性定义索引库的一些基础配置。
input { stdin { }}filter { }output { elasticsearch { hosts => "192.168.80.10:9200" }}
继续
Filter plugins
本文转自大数据躺过的坑博客园博客,原文链接:http://www.cnblogs.com/zlslch/p/6624457.html,如需转载请自行联系原作者