网站的关键字 设置/史上最强大的搜索神器

第3章 Logstash数据分析

Logstash使用管道方式进行日志的搜集处理和输出。有点类似*NIX系统的管道命令 xxx | ccc | ddd，xxx执行完了会执行ccc，然后执行ddd。
在logstash中，包括了三个阶段:
输入input --> 处理filter（不是必须的） --> 输出output

每个阶段都由很多的插件配合工作，比如file、elasticsearch、redis等等。
每个阶段也可以指定多种方式，比如输出既可以输出到elasticsearch中，也可以指定到stdout在控制台打印。

logstash支持多输入和多输出

ELFK架构示意图：

1.Logstash基础部署

1. 安装软件

[root@host3 ~]# yum install logstash --enablerepo=es -y 			# 偶尔需要使用的仓库可以将它关闭，用到的时候临时打开[root@host3 ~]# ln -sv /usr/share/logstash/bin/logstash /usr/local/bin/	# 做软连接，命令就可以直接使用了
"/usr/local/bin/logstash" -> "/usr/share/logstash/bin/logstash"

2. 创建第一个配置文件

[root@host3 ~]# vim 01-stdin-stdout.confinput {stdin {}
}output {stdout {}
}

3. 测试配置文件

[root@host3 ~]# logstash -tf 01-stdin-stdout.conf 

4. 自定义启动，这种方式通常用于实验环境，业务环境下，通常将配置修改后，使用systemctl来管理服务

[root@host3 ~]# logstash -f 01-stdin-stdout.conf 
Using bundled JDK: /usr/share/logstash/jdk
OpenJDK 64-Bit Server VM warning: Option UseConcMarkSweepGC was deprecated in version 9.0 and will likely be removed in a future release.
WARNING: Could not find logstash.yml which is typically located in $LS_HOME/config or /etc/logstash. You can specify the path using --path.settings. Continuing using the defaults
Could not find log4j2 configuration at path /usr/share/logstash/config/log4j2.properties. Using default config which logs errors to the console
[INFO ] 2022-09-15 21:49:37.109 [main] runner - Starting Logstash {"logstash.version"=>"7.17.6", "jruby.version"=>"jruby 9.2.20.1 (2.5.8) 2021-11-30 2a2962fbd1 OpenJDK 64-Bit Server VM 11.0.16+8 on 11.0.16+8 +indy +jit [linux-x86_64]"}
[INFO ] 2022-09-15 21:49:37.115 [main] runner - JVM bootstrap flags: [-Xms1g, -Xmx1g, -XX:+UseConcMarkSweepGC, -XX:CMSInitiatingOccupancyFraction=75, -XX:+UseCMSInitiatingOccupancyOnly, -Djava.awt.headless=true, -Dfile.encoding=UTF-8, -Djdk.io.File.enableADS=true, -Djruby.compile.invokedynamic=true, -Djruby.jit.threshold=0, -Djruby.regexp.interruptible=true, -XX:+HeapDumpOnOutOfMemoryError, -Djava.security.egd=file:/dev/urandom, -Dlog4j2.isThreadContextMapInheritable=true]
[INFO ] 2022-09-15 21:49:37.160 [main] settings - Creating directory {:setting=>"path.queue", :path=>"/usr/share/logstash/data/queue"}
[INFO ] 2022-09-15 21:49:37.174 [main] settings - Creating directory {:setting=>"path.dead_letter_queue", :path=>"/usr/share/logstash/data/dead_letter_queue"}
[WARN ] 2022-09-15 21:49:37.687 [LogStash::Runner] multilocal - Ignoring the 'pipelines.yml' file because modules or command line options are specified
[INFO ] 2022-09-15 21:49:38.843 [LogStash::Runner] Reflections - Reflections took 114 ms to scan 1 urls, producing 119 keys and 419 values 
[WARN ] 2022-09-15 21:49:39.658 [LogStash::Runner] line - Relying on default value of `pipeline.ecs_compatibility`, which may change in a future major release of Logstash. To avoid unexpected changes when upgrading Logstash, please explicitly declare your desired ECS Compatibility mode.
[WARN ] 2022-09-15 21:49:39.703 [LogStash::Runner] stdin - Relying on default value of `pipeline.ecs_compatibility`, which may change in a future major release of Logstash. To avoid unexpected changes when upgrading Logstash, please explicitly declare your desired ECS Compatibility mode.
Configuration OK
[INFO ] 2022-09-15 21:49:39.917 [LogStash::Runner] runner - Using config.test_and_exit mode. Config Validation Result: OK. Exiting Logstash
[root@host3 ~]# logstash -f 01-stdin-stdout.conf 
Using bundled JDK: /usr/share/logstash/jdk
OpenJDK 64-Bit Server VM warning: Option UseConcMarkSweepGC was deprecated in version 9.0 and will likely be removed in a future release.WARNING: Could not find logstash.yml which is typically located in $LS_HOME/config or /etc/logstash. You can specify the path using --path.settings. Continuing using the defaults
Could not find log4j2 configuration at path /usr/share/logstash/config/log4j2.properties. Using default config which logs errors to the console
[INFO ] 2022-09-15 21:50:25.095 [main] runner - Starting Logstash {"logstash.version"=>"7.17.6", "jruby.version"=>"jruby 9.2.20.1 (2.5.8) 2021-11-30 2a2962fbd1 OpenJDK 64-Bit Server VM 11.0.16+8 on 11.0.16+8 +indy +jit [linux-x86_64]"}
[INFO ] 2022-09-15 21:50:25.103 [main] runner - JVM bootstrap flags: [-Xms1g, -Xmx1g, -XX:+UseConcMarkSweepGC, -XX:CMSInitiatingOccupancyFraction=75, -XX:+UseCMSInitiatingOccupancyOnly, -Djava.awt.headless=true, -Dfile.encoding=UTF-8, -Djdk.io.File.enableADS=true, -Djruby.compile.invokedynamic=true, -Djruby.jit.threshold=0, -Djruby.regexp.interruptible=true, -XX:+HeapDumpOnOutOfMemoryError, -Djava.security.egd=file:/dev/urandom, -Dlog4j2.isThreadContextMapInheritable=true]
[WARN ] 2022-09-15 21:50:25.523 [LogStash::Runner] multilocal - Ignoring the 'pipelines.yml' file because modules or command line options are specified
[INFO ] 2022-09-15 21:50:25.555 [LogStash::Runner] agent - No persistent UUID file found. Generating new UUID {:uuid=>"3fc04af1-7665-466e-839f-1eb42348aeb0", :path=>"/usr/share/logstash/data/uuid"}
[INFO ] 2022-09-15 21:50:27.119 [Api Webserver] agent - Successfully started Logstash API endpoint {:port=>9600, :ssl_enabled=>false}
[INFO ] 2022-09-15 21:50:28.262 [Converge PipelineAction::Create<main>] Reflections - Reflections took 110 ms to scan 1 urls, producing 119 keys and 419 values 
[WARN ] 2022-09-15 21:50:29.084 [Converge PipelineAction::Create<main>] line - Relying on default value of `pipeline.ecs_compatibility`, which may change in a future major release of Logstash. To avoid unexpected changes when upgrading Logstash, please explicitly declare your desired ECS Compatibility mode.
[WARN ] 2022-09-15 21:50:29.119 [Converge PipelineAction::Create<main>] stdin - Relying on default value of `pipeline.ecs_compatibility`, which may change in a future major release of Logstash. To avoid unexpected changes when upgrading Logstash, please explicitly declare your desired ECS Compatibility mode.
[INFO ] 2022-09-15 21:50:29.571 [[main]-pipeline-manager] javapipeline - Starting pipeline {:pipeline_id=>"main", "pipeline.workers"=>2, "pipeline.batch.size"=>125, "pipeline.batch.delay"=>50, "pipeline.max_inflight"=>250, "pipeline.sources"=>["/root/01-stdin-stdout.conf"], :thread=>"#<Thread:0x32e464e6 run>"}
[INFO ] 2022-09-15 21:50:30.906 [[main]-pipeline-manager] javapipeline - Pipeline Java execution initialization time {"seconds"=>1.33}
WARNING: An illegal reflective access operation has occurred
WARNING: Illegal reflective access by com.jrubystdinchannel.StdinChannelLibrary$Reader (file:/usr/share/logstash/vendor/bundle/jruby/2.5.0/gems/jruby-stdin-channel-0.2.0-java/lib/jruby_stdin_channel/jruby_stdin_channel.jar) to field java.io.FilterInputStream.in
WARNING: Please consider reporting this to the maintainers of com.jrubystdinchannel.StdinChannelLibrary$Reader
WARNING: Use --illegal-access=warn to enable warnings of further illegal reflective access operations
WARNING: All illegal access operations will be denied in a future release
[INFO ] 2022-09-15 21:50:31.128 [[main]-pipeline-manager] javapipeline - Pipeline started {"pipeline.id"=>"main"}
The stdin plugin is now waiting for input:
[INFO ] 2022-09-15 21:50:31.270 [Agent thread] agent - Pipelines running {:count=>1, :running_pipelines=>[:main], :non_running_pipelines=>[]}
abc
{"message" => " abc","@version" => "1","host" => "host3.test.com","@timestamp" => 2022-09-15T13:52:02.984Z
}
bbb
{"message" => "bbb","@version" => "1","host" => "host3.test.com","@timestamp" => 2022-09-15T13:52:06.177Z
}

2.输入类型

在上例中，输入类型是stdin，也就是手动输入，而在生产环境中，日志不可能通过手工输入的发生产生，因此stdin通常都是用于测试环境是否搭建成功，下面会介绍几种常见的输入类型。

2.1 file

input {file {path => ["/tmp/test/*.txt"]# 从最开始读日志文件（默认是末尾），仅在读取记录没有任何记录的情况下生效，也就是说，在服务停止的时候有新文件产生，服务器启动后可以读取到（旧文件不行）start_position => "beginning"  }
}

文件的读取记录放在/usr/share/logstash/data/plugins/inputs/file/.sincedb_3cd99a80ca58225ec14dc0ac340abb80中

[root@host3 ~]# cat /usr/share/logstash/data/plugins/inputs/file/.sincedb_3cd99a80ca58225ec14dc0ac340abb80
5874000 0 64768 4 1663254379.147252 /tmp/test/1.txt

2.2 tcp

和filebeat一样，Logstash同样支持监听TCP的某一个端口，用来接收日志。可以同时监听多个端口

这种方式通常用于无法安装客户端的服务器

也可以使用HTTP协议，配置方法和TCP类似

[root@host3 ~]#vim 03-tcp-stdout.conf 
input {tcp {port => 9999}
}output {stdout {}
}

[root@host2 ~]# telnet 192.168.19.103 9999
Trying 192.168.19.103...
Connected to 192.168.19.103.
Escape character is '^]'.
123456
test
hello

{"message" => "123456\r","@version" => "1","@timestamp" => 2022-09-15T15:30:23.123Z,"host" => "host2","port" => 51958
}
{"message" => "test\r","@version" => "1","@timestamp" => 2022-09-15T15:30:24.494Z,"host" => "host2","port" => 51958
}
{"message" => "hello\r","@version" => "1","@timestamp" => 2022-09-15T15:30:26.336Z,"host" => "host2","port" => 51958
}

2.3 redis

Logstash支持直接从redis数据库中拿数据。支持三种redis数据类型：

1. list，表示的redis命令为blpop，代表从redis list的左边获取第一个元素，如无元素则阻塞；
2. channel，表示的redis命令为subscribe，代表从redis频道获取最新的数据；
3. pattern_channel，表示的redis命令为psubscribe，代表通过pattern正则表达式匹配频道，获取最新的数据。

数据类型之间的区别：

1. channel与pattern_channel的区别在于，pattern_channel可以通过正则表达式匹配多个频道，而channel是单一频道；
2. list与另外两个channel的区别在于，1个channel的数据会被多个订阅的logstash重复获取，1个list的数据被多个logstash获取时不会重复，会被分摊在各个Logstash中。

输入配置如下

input { redis {data_type => "list"         # 指定数据类型db => 5                     # 指定数据库，默认是0host => "192.168.19.101"    # 指定redis服务器IP，默认是localhostport => 6379password => "bruce"key => "test-list"}
}

redis中追加数据

[root@host1 ~]# redis-cli -h host1 -a bruce
host1:6379> select 5
OK
host1:6379[5]> lpush test-list bruce
(integer) 1
host1:6379[5]> lrange test-list 0 -1
(empty list or set)
host1:6379[5]> lpush test-list hello
(integer) 1
host1:6379[5]> lrange test-list 0 -1		# 可以看到，Logstash获取数据后，会将列表清空
(empty list or set)
host1:6379[5]> lpush test-list '{"requestTime":"[12/Sep/2022:23:30:56 +0800]","clientIP":"192.168.19.1","threadID":"http-bio-8080-exec-7","protocol":"HTTP/1.1","requestMethod":"GET / HTTP/1.1","requestStatus":"404","sendBytes":"-","queryString":"","responseTime":"0ms","partner":"-","agentVersion":"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/105.0.0.0 Safari/537.36"}'

Logstash获取数据

{"message" => "bruce","@timestamp" => 2022-09-16T08:17:38.213Z,"@version" => "1","tags" => [[0] "_jsonparsefailure"]
}
# 非json格式数据会报错，但是能接收
[ERROR] 2022-09-16 16:18:21.688 [[main]<redis] json - JSON parse error, original data now in message field {:message=>"Unrecognized token 'hello': was expecting ('true', 'false' or 'null')\n at [Source: (String)\"hello\"; line: 1, column: 11]", :exception=>LogStash::Json::ParserError, :data=>"hello"}
{"message" => "hello","@timestamp" => 2022-09-16T08:18:21.689Z,"@version" => "1","tags" => [[0] "_jsonparsefailure"]
}
# json格式的数据过来，Logstash可以自动解析
{"clientIP" => "192.168.19.1","requestTime" => "[12/Sep/2022:23:30:56 +0800]","queryString" => "","@version" => "1","agentVersion" => "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/105.0.0.0 Safari/537.36","partner" => "-","@timestamp" => 2022-09-16T08:23:10.320Z,"protocol" => "HTTP/1.1","requestStatus" => "404","threadID" => "http-bio-8080-exec-7","requestMethod" => "GET / HTTP/1.1","sendBytes" => "-","responseTime" => "0ms"
}

2.4 beats

在FileBeat中已经配置好了将日志输出到Logstash，在Logstash中，只需要接收数据即可。

filebeat配置

filebeat.inputs:
- type: logpaths: /tmp/1.txtoutput.logstash:hosts: ["192.168.19.103:5044"]

Logstash配置

input { beats {port => 5044}
}

host2上在/tmp/1.txt中追加111，Logstash的输出

{"message" => "111","tags" => [[0] "beats_input_codec_plain_applied"],"agent" => {"id" => "76b7876b-051a-4df8-8b13-bd013ac5ec59","version" => "7.17.4","hostname" => "host2.test.com","type" => "filebeat","name" => "host2.test.com","ephemeral_id" => "437ac89f-7dc3-4898-a457-b2452ac4223b"},"input" => {"type" => "log"},"host" => {"name" => "host2.test.com"},"log" => {"offset" => 0,"file" => {"path" => "/tmp/1.txt"}},"@version" => "1","ecs" => {"version" => "1.12.0"},"@timestamp" => 2022-09-16T08:53:20.975Z
}

3. 输出类型

3.1 redis

redis也可以作为输出类型，配置方式和输入类似

output { redis {data_type => "list" db => 6 host => "192.168.19.101" port => 6379password => "bruce"key => "test-list"}
}

查看redis数据库

[root@host1 ~]# redis-cli -h host1 -a bruce
host1:6379> select 6
OK
host1:6379[6]> lrange test-list 0 -1
1) "{\"message\":\"1111\",\"@version\":\"1\",\"@timestamp\":\"2022-09-16T09:12:29.890Z\",\"host\":\"host3.test.com\"}"

3.2 file

file类型是输出到本地磁盘保存。

output { file {path => "/tmp/test-file.log"}
}

3.3 elasticsearch

output { elasticsearch {hosts => ["192.168.19.101:9200","192.168.19.102:9200","192.168.19.103:9200"]index => "centos-logstash-elasticsearh-%{+YYYY.MM.dd}"}
}

4. filter

filter是一个可选插件，在接收到日志信息后，可以对日志进行格式化，然后再输出。

4.1 grok

grok可以用来解析任意文本并进行结构化。该工具适合syslog日志、Apache和其他网络服务器日志。

①简单示例

---

input {file {path => ["/var/log/nginx/access.log*"]start_position => "beginning"}
}filter {grok {match => {"message" => "%{COMBINEDAPACHELOG}"# "message" => "%{HTTPD_COMMONLOG}"		# 新版本Logstash可能会用这个变量}}
}output {stdout {}elasticsearch {hosts => ["192.168.19.101:9200","192.168.19.102:9200","192.168.19.103:9200"]index => "nginx-logs-es-%{+YYYY.MM.dd}"}
}

解析出来的结果：

{"request" => "/","bytes" => "4833","@version" => "1","auth" => "-","agent" => "\"curl/7.29.0\"","path" => "/var/log/nginx/access.log-20220913","ident" => "-","verb" => "GET","message" => "192.168.19.102 - - [12/Sep/2022:21:48:29 +0800] \"GET / HTTP/1.1\" 200 4833 \"-\" \"curl/7.29.0\" \"-\"","httpversion" => "1.1","host" => "host3.test.com","@timestamp" => 2022-09-16T14:27:43.208Z,"response" => "200","timestamp" => "12/Sep/2022:21:48:29 +0800","referrer" => "\"-\"","clientip" => "192.168.19.102"
}

②预定义字段

---

grok是基于正则表达式来进行匹配，它的语法格式是%{SYNTAX:SEMANTIC}

• SYNTAX是将匹配您的文本的模式的名称，这是内置好的语法，官方支持120种字段。
• SEMANTIC是您为要匹配的文本提供的标识符，也就是你要给它去的名字。

示例：

1. 日志源文件

55.3.244.1 GET /index.html 15824 0.043

2. 匹配的字段应该是

    %{IP:client} %{WORD:method} %{URIPATHPARAM:request} %{NUMBER:bytes} %{NUMBER:duration}

3. 配置文件

input {stdin {}
}filter {grok {match => { "message" => "%{IP:client} %{WORD:method} %{URIPATHPARAM:request} %{NUMBER:bytes} %{NUMBER:duration}" }}
}output {stdout {}
}

4. 匹配出来的结果

55.3.244.1 GET /index.html 15824 0.043
{"message" => "55.3.244.1 GET /index.html 15824 0.043","@version" => "1","@timestamp" => 2022-09-16T14:46:46.426Z,"method" => "GET","request" => "/index.html","bytes" => "15824","duration" => "0.043","host" => "host3.test.com","client" => "55.3.244.1"
}

针对不同服务的日志，可以查看官方文档的定义：
https://github.com/logstash-plugins/logstash-patterns-core/tree/master/patterns

③自定义字段

---

当预定义的字段不符合要求时，grok也支持自定义正则表达式来匹配日志信息

1. 首先需要创建自定义表达式保存的目录，并将表达式写进去

[root@host3 ~]# mkdir patterns
[root@host3 ~]# echo "POSTFIX_QUEUEID [0-9A-F]{10,11}" >> ./patterns/1

2. 修改配置文件

input {stdin {}
}filter {grok {patterns_dir => ["/root/patterns"]											# 指定表达式位置match => { "message" => "%{SYSLOGBASE} %{POSTFIX_QUEUEID:queue_id}: %{GREEDYDATA:syslog_message}" }	# 这里有系统预定义的，也有自定义的表达式，大括号外的字符就是常规的字符，需要逐个匹配，如冒号: }
}output {stdout {}
}

3. 运行并测试

...
The stdin plugin is now waiting for input:
[INFO ] 2022-09-16 23:22:04.511 [Agent thread] agent - Pipelines running {:count=>1, :running_pipelines=>[:main], :non_running_pipelines=>[]}
Jan  1 06:25:43 mailserver14 postfix/cleanup[21403]: BEF25A72965: message-id=<[e-mail]>
{"message" => "Jan  1 06:25:43 mailserver14 postfix/cleanup[21403]: BEF25A72965: message-id=<[e-mail]>","host" => "host3.test.com","timestamp" => "Jan  1 06:25:43","queue_id" => "BEF25A72965",			# 自定义表达式匹配的字段"logsource" => "mailserver14","@timestamp" => 2022-09-16T15:22:19.516Z,"program" => "postfix/cleanup","pid" => "21403","@version" => "1","syslog_message" => "message-id=<[e-mail]>"
}

4.2 通用字段

顾名思义，这些字段可以用在所有属于filter的插件中。

• remove_field

filter {grok {remove_field => ["@version","tag","agent"]}
}

• add_field

filter {grok {add_field => ["new_tag" => "hello world %{YYYY.mm.dd}"]}
}

4.3 date

在数据中，会有两个时间戳timestamp和@timestamp，日志产生的时间和数据采集的时间，这两个时间可能会不一致。
date插件可以用来转换日志记录中的时间字符串，参考@timestamp字段里的时间。date插件支持五种时间格式：

• ISO8601
• UNIX
• UNIX_MS
• TAI64N

input {file {path => "/var/log/nginx/access.log*"start_position => "beginning"}
}filter {grok {match => { "message" => "%{HTTPD_COMMONLOG}" }remove_field => ["message","ident","auth","@version","path"]}date {match => [ "timestamp","dd/MMM/yyyy:HH:mm:ss Z" ]		# timestamp必须是现有的字段，这里只是对这个字段的时间进行校正，且需要和timestamp字段的原数据格式一致，否则会报解析错误# timestamp原来的数据格式为"17/Sep/2022:18:42:26 +0800"，因此时区改成ZZZ就会一直报错，因为ZZZ代表Asia/Shanghai这种格式，Z代表+0800timezone => "Asia/Shanghai"}}output {stdout {}
}

输出的格式：

{"timestamp" => "17/Sep/2022:18:42:26 +0800", #和@timestamp有8小时的时间差，可到Elasticsearch中查看，如果也有时间差，可以在date中修改timezone"response" => "200","httpversion" => "1.1","clientip" => "192.168.19.102","verb" => "GET","host" => "host3.test.com","request" => "/","@timestamp" => 2022-09-17T10:42:26.000Z,"bytes" => "4833"
}

使用target将匹配到的时间字段解析后存储到目标字段，若不指定，默认是@timestamp字段。这个字段在Kibana中创建索引时可以用到

  date {match => [ "timestamp","dd/MMM/yyyy:HH:mm:ss Z" ]timezone => "Asia/Shanghai"target => "logtime"}# 结果
{"timestamp" => "17/Sep/2022:21:15:30 +0800","response" => "200","logtime" => 2022-09-17T13:15:30.000Z,		# 日志产生的时间"httpversion" => "1.1","clientip" => "192.168.19.102","verb" => "GET","host" => "host3.test.com","request" => "/","@timestamp" => 2022-09-17T13:15:31.357Z,		# 日志记录的时间，可以看到和日志产生的时间有一定的延迟"bytes" => "4833"
}

4.4 geoip

用来解析访问IP的位置信息。这个插件是依赖GeoLite2城市数据库，信息不一定准确，也可以自己下载MaxMind格式的数据库然后应用，官方网站有自定义数据库的指导手册。

input {file {path => "/var/log/nginx/access.log*"start_position => "beginning"}
}filter {grok {match => { "message" => "%{HTTPD_COMMONLOG}" }remove_field => ["message","ident","auth","@version","path"]}geoip {source => "clientip" 			# IP地址的源参考clientip字段# fields => ["country_name" ,"timezone", "city_name"]		# 可以选择显示的字段}}output {stdout {}
}

得到的结果，可以看到，私有地址无法正常解析

{"timestamp" => "17/Sep/2022:21:15:30 +0800","response" => "200","geoip" => {},"httpversion" => "1.1","clientip" => "192.168.19.102","verb" => "GET","host" => "host3.test.com","tags" => [[0] "_geoip_lookup_failure"				# 私网地址],"request" => "/","@timestamp" => 2022-09-17T13:30:05.178Z,"bytes" => "4833"
}
{"timestamp" => "17/Sep/2022:21:15:30 +0800","response" => "200","geoip" => {					# 解析的结果放在geoip中"country_code2" => "CM","country_code3" => "CM","country_name" => "Cameroon","ip" => "154.72.162.134","timezone" => "Africa/Douala","location" => {"lon" => 12.5,"lat" => 6.0},"continent_code" => "AF","latitude" => 6.0,"longitude" => 12.5},"httpversion" => "1.1","clientip" => "154.72.162.134","verb" => "GET","host" => "host3.test.com","request" => "/","@timestamp" => 2022-09-17T13:30:05.178Z,"bytes" => "4833"
}

4.5 useragent

用来解析浏览器的信息。前提是输出的信息有浏览器信息字段。

input {file {path => "/var/log/nginx/access.log*"start_position => "beginning"}
}filter {grok {match => { "message" => "%{HTTPD_COMBINEDLOG}" }		# HTTPD_COMBINEDLOG可以解析浏览器remove_field => ["message","ident","auth","@version","path"]}useragent {source => "agent"							# 指定浏览器信息在哪个字段中，这个字段必须要存在target => "agent_test"						# 为了方便查看，将所有解析后的信息放到这个字段里面去}
}output {stdout {}
}

得到的结果：

{"timestamp" => "17/Sep/2022:23:42:31 +0800","response" => "404","geoip" => {},"httpversion" => "1.1","clientip" => "192.168.19.103","verb" => "GET","agent" => "\"Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Firefox/60.0\"","host" => "host3.test.com","request" => "/favicon.ico","referrer" => "\"-\"","@timestamp" => 2022-09-17T15:42:31.927Z,"bytes" => "3650","agent_test" => {"major" => "60","name" => "Firefox","os" => "Linux","os_full" => "Linux","os_name" => "Linux","version" => "60.0","minor" => "0","device" => "Other"}
}
{
{
..."agent_test" => {"major" => "60","name" => "Firefox","os" => "Linux","os_full" => "Linux","os_name" => "Linux","version" => "60.0","minor" => "0","device" => "Other"}
}
{
..."agent_test" => {"os_minor" => "0","os_full" => "iOS 16.0","version" => "16.0","os_major" => "16","device" => "iPhone","major" => "16","name" => "Mobile Safari","os" => "iOS","os_version" => "16.0","os_name" => "iOS","minor" => "0"}
}
{
..."agent_test" => {"patch" => "3987","os_full" => "Android 10","version" => "80.0.3987.162","os_major" => "10","device" => "Samsung SM-G981B","major" => "80","name" => "Chrome Mobile","os" => "Android","os_version" => "10","os_name" => "Android","minor" => "0"}
}

4.6 mutate

1. 切割自定的字段

input {stdin {}
}filter {mutate {split => {message =>  " "		# 将message消息以空格作为分隔符进行分割}remove_field => ["@version","host"]add_field => {"tag" => "This a test field from Bruce"}}
}output {stdout {}
}

111 222 333
{"tag" => "This a test field from Bruce","message" => [[0] "111",[1] "222",[2] "333"],"@timestamp" => 2022-09-18T08:07:36.373Z
}

2. 将切割后的数据取出来

input {stdin {}
}filter {mutate {split => {message =>  " "		# 将message消息以空格作为分隔符进行分割}remove_field => ["@version","host"]add_field => {"tag" => "This a test field from Bruce"}}mutate {add_field => {"name" => "%{[message][0]}""age" => "%{[message][1]}""sex" => "%{[message][2]}"}}
}output {stdout {}
}

bruce 37 male
{"message" => [[0] "bruce",[1] "37",[2] "male"],"age" => "37","@timestamp" => 2022-09-18T08:14:31.230Z,"sex" => "male","tag" => "This a test field from Bruce","name" => "bruce"
}

3. convert：将字段的值转换成不同的类型，例如将字符串转换成证书，如字段值是一个数组，所有成员都会被转换。如果该字段是散列，则不会采取任何动作

filter {mutate {convert => {"age" => "integer"			# 将age转换成数字类型}}
}

bruce 20 male
{"message" => [[0] "bruce",[1] "20",[2] "male"],"sex" => "male","name" => "bruce","age" => 20,					# 没有引号，代表已经修改成数字类型了"@timestamp" => 2022-09-18T08:51:07.633Z,"tag" => "This a test field from Bruce"
}

4. strip：剔除字段中的前导和尾随的空格

filter {mutate {strip => { "name","sex" }}
}

5. rename：修改字段名

filter {mutate {rename => { "sex" => "agenda" }}
}

6. replace：替换字段内容

filter {mutate {replace => { "tag" => "This is test message" }		# 修改了tag字段的内容}
}

7. update：用法和replace一样，区别在于如果字段存在则修改内容，如果过不存在则忽略此操作

8. uppercase/lowercase：转换成大写/小写；capitalize：首字母大写。转换的是字段内容

filter {mutate {uppercase => "tag" capitalize => "name" }
}

5 高级特性

5.1 判断语法

在input中打上标记后，可以在output和filter中通过判断语句来做区别化的处理

input {beats {port => 8888type => "nginx-beats"}tcp {port => 9999type => "tomcat-tcp"}
}output { if [type] == "nginx-beats" {elasticsearch {hosts => ["192.168.19.101:9200","192.168.19.102:9200","192.168.19.103:9200"]index => "nginx-beats-elasticsearh-%{+YYYY.MM.dd}"}} else {elasticsearch {hosts => ["192.168.19.101:9200","192.168.19.102:9200","192.168.19.103:9200"]index => "tomcat-tcp-elasticsearh-%{+YYYY.MM.dd}"}
}

5.2 多实例运行

Logstash支持多实例运行，但是如果直接启动，第二个实例会报错，需要指定path.data的路径才能正常启动。

[root@host3 ~]# logstash -f 01-stdin-stdout.conf --path.data /tmp/logstash