大数据入门第二十五天——logstash入门
一、概述
1.logstash是什么
根据官网介绍:
Logstash 是开源的服务器端数据处理管道,能够同时 从多个来源采集数据、转换数据,然后将数据发送到您最喜欢的 “存储库” 中。(我们的存储库当然是 Elasticsearch。)
//属于elasticsearch旗下产品(JRuby开发,开发者曾说如果他知道有scala,就不会用jruby了。。)
也就是说,它是flume的“后浪”,它解决了“前浪”flume的数据丢失等问题!
2.基础结构
输入:采集各种来源数据
过滤:实时解析转换数据
输出:选择存储库导出数据
补充:Logstash 每读取一次数据的行为叫做事件。
更多详细介绍,包括具体支持的输入输出等,参考:https://www.elastic.co/guide/index.html
用法博文推荐:https://blog.csdn.net/chenleiking/article/details/73563930
二、安装
logstash5.x 6.x需要JDK1.8+,如未安装,请先安装JDK1.8+
1.下载
https://www.elastic.co/downloads/past-releases
选择合适的版本,下载即可
2.解压
[hadoop@mini1 ~]$ tar -zxvf logstash-5.6.9.tar.gz -C apps/
三、入门使用
1.HelloWorld示例
运行启动命令,并直接给出配置
bin/logstash -e 'input { stdin { } } output { stdout {} }'
常用的启动参数如下:
运行结果如下:输入helloworld,给出message消息:
[hadoop@mini1 logstash-5.6.9]$ bin/logstash -e 'input { stdin { } } output { stdout {} }'
Sending Logstash's logs to /home/hadoop/apps/logstash-5.6.9/logs which is now configured via log4j2.properties
[2018-04-18T16:38:18,309][INFO ][logstash.modules.scaffold] Initializing module {:module_name=>"fb_apache", :directory=>"/home/hadoop/apps/logstash-5.6.9/modules/fb_apache/configuration"}
[2018-04-18T16:38:18,323][INFO ][logstash.modules.scaffold] Initializing module {:module_name=>"netflow", :directory=>"/home/hadoop/apps/logstash-5.6.9/modules/netflow/configuration"}
[2018-04-18T16:38:18,325][INFO ][logstash.setting.writabledirectory] Creating directory {:setting=>"path.queue", :path=>"/home/hadoop/apps/logstash-5.6.9/data/queue"}
[2018-04-18T16:38:18,373][INFO ][logstash.setting.writabledirectory] Creating directory {:setting=>"path.dead_letter_queue", :path=>"/home/hadoop/apps/logstash-5.6.9/data/dead_letter_queue"}
[2018-04-18T16:38:18,401][INFO ][logstash.agent ] No persistent UUID file found. Generating new UUID {:uuid=>"893c481c-85d1-4746-8562-48a74dcbad08", :path=>"/home/hadoop/apps/logstash-5.6.9/data/uuid"}
[2018-04-18T16:38:18,757][INFO ][logstash.pipeline ] Starting pipeline {"id"=>"main", "pipeline.workers"=>1, "pipeline.batch.size"=>125, "pipeline.batch.delay"=>5, "pipeline.max_inflight"=>125}
[2018-04-18T16:38:18,847][INFO ][logstash.pipeline ] Pipeline main started
The stdin plugin is now waiting for input:
[2018-04-18T16:38:18,929][INFO ][logstash.agent ] Successfully started Logstash API endpoint {:port=>9600}
HelloWorld
{
"@version" => "1",
"host" => "mini1",
"@timestamp" => 2018-04-18T08:38:44.798Z,
"message" => "HelloWorld"
}
2.使用配置文件
实际中的 -e 后的配置一般相对更复杂,所以一般会通过 -f 使用配置文件来启动
bin/logstash -f logstash.conf
配置文件大概长这样:
# 输入
input {
...
}
# 过滤器
filter {
...
}
# 输出
output {
...
}
编写一个示例的配置文件:logstash.conf:
input {
# 从文件读取日志信息
file {
path => "/home/hadoop/apps/logstash-5.6.9/logs/1.log"
type => "system"
start_position => "beginning"
}
}
# filter {
#
# }
output {
# 标准输出
stdout { codec => rubydebug }
}
输出结果如下:
[hadoop@mini1 logstash-5.6.9]$ bin/logstash -f logstash.conf
Sending Logstash's logs to /home/hadoop/apps/logstash-5.6.9/logs which is now configured via log4j2.properties
[2018-04-18T16:49:28,522][INFO ][logstash.modules.scaffold] Initializing module {:module_name=>"fb_apache", :directory=>"/home/hadoop/apps/logstash-5.6.9/modules/fb_apache/configuration"}
[2018-04-18T16:49:28,525][INFO ][logstash.modules.scaffold] Initializing module {:module_name=>"netflow", :directory=>"/home/hadoop/apps/logstash-5.6.9/modules/netflow/configuration"}
[2018-04-18T16:49:28,988][INFO ][logstash.pipeline ] Starting pipeline {"id"=>"main", "pipeline.workers"=>1, "pipeline.batch.size"=>125, "pipeline.batch.delay"=>5, "pipeline.max_inflight"=>125}
[2018-04-18T16:49:29,371][INFO ][logstash.pipeline ] Pipeline main started
{
"@version" => "1",
"host" => "mini1",
"path" => "/home/hadoop/apps/logstash-5.6.9/logs/1.log",
"@timestamp" => 2018-04-18T08:49:29.451Z,
"message" => "Apr 16 17:01:01 mini1 systemd: Started Session 5 of user root.",
"type" => "system"
}
四、插件的使用
logstash主要有3个主插件:输入input,输出output,过滤filter,其他还包括编码解码插件等
1.输入插件input
定义的数据源,支持从文件、stdin、kafka、twitter等来源,甚至可以自己写一个input plugin。
输入的file path等是支持通配的,例如:
path => "/data/web/logstash/logFile/*/*.log"
常用输入插件
1.file
file插件的必选参数只有path一项:部分选项如下:
// 原版的完整参数解释参见官网,中文参见上文博文参考处链接
配置示例:
input
file {
path => ["/var/log/*.log", "/var/log/message"]
type => "system"
start_position => "beginning"
}
}
2.过滤插件、输出插件
同输入插件类似,可以参考官网详细配置与参考博文