ELK安装、部署、调试 （八）logstash配置语法详解

input {
  #输入插件
}

filter {
  #过滤插件
}

output {
  #输出插件
}

1.读取文件。

使用filewatch的ruby gem库来监听文件变化，并通过.sincedb的数据库文件记录被监听日志we年的读取进度（时间

搓）

。sincedb数据文件的默认路径为<path.data>/plugins/inputs/file下面，文件名类似

于.sincedb_234534534sdfgsfd23，<path.data>为logstash的插件存储目录默认是LOGSTASH_HOME/data实验一：本机/var/log/secure为输入日志，标准输出

vi /usr/local/logstash/2logstash-1.conf

input {
  file {
    path => ["/var/log/messages"],[]
    type => "ly_system"
    start_position => "beginning"       
#从beginning也就是文件开头进行读取，如果不写，默认是从文件最后开始读取。                                
        #如果不想把文件全部作为输入，就不配置此属性。
  }
}

output {
  stdout {
    codec => rubydebug

  }
}

1.保存10.10.10.74 f-kafka-logs-es.conf的配置信息

[root@localhost logstash]# cat f-kafka-logs-es.conf
input {
    kafka {
    bootstrap_servers => "10.10.10.71:9092,10.10.10.72:9092,10.10.10.73:9092"
    topics => ["osmessages"]
    }
}
output {
    elasticsearch {
    hosts => ["10.10.10.65:9200","10.10.10.66:9200","10.10.10.67:9200"]
    index => "osmessageslog-%{+YYYY-MM-dd}"

  }
}

2.停止logstash服务

kill -9 13508

3.使用2logstash-1.conf作为配置文件启动logstash

nohup /usr/local/logstash/bin/logstash -f /usr/local/logstash/2logstash-1.conf &4.查看

tail -f nohup.out

{
          "type" => "ly_system",
          "path" => "/var/log/secure",
      "@version" => "1",
          "host" => "localhost.localdomain",
       "message" => "Aug 31 08:43:56 localhost sshd[6920]: Accepted password for root from
172.16.17.234 port 1909 ssh2",
    "@timestamp" => 2023-08-31T00:52:13.054Z
}

实验二：input插件添加域和标签

[root@localhost logstash]# cat 2logstash-1.conf
input {
  file {
    path => ["/var/log/secure"]
    type => "ly_system"
    start_position => "beginning"
    add_field => {"I'm " => "10.10.10.74"}
    tags => ["74","logstash1"]

#从beginning也就是文件开头进行读取，如果不写，默认是从文件最后开始读取。

#如果不想把文件全部作为输入，就不配置此属性。

  }
}
output {
  stdout {
    codec => rubydebug

  }
}

***************************

add_field => {"I'm " => "10.10.10.74"} 添加一个新的域，自己定义的。

tags => "74","logstash1" tags是内置的域，可以用来定义标签。

***************************

输出结果

{
      "@version" => "1",
          "host" => "localhost.localdomain",
          "I'm " => "10.10.10.74",
          "path" => "/var/log/secure",
       "message" => "Aug 31 09:03:33 localhost sshd[14339]: Accepted password for root from
172.16.17.234 port 2684 ssh2",
    "@timestamp" => 2023-08-31T01:03:53.586Z,
          "type" => "ly_system",
          "tags" => [
        [0] "74",
        [1] "logstash1"
    ]
}

实验三：input读取syslog日志。

需要完成2个步骤的操作，

1，vi /etc/rsyslog.conf

*.* @@10.10.10.74:5514 #10.10.10.74本机logstash服务器的IP地址，这个配置时使用rsyslog客户端把本机

的日志信息传输到10.10.10.74服务器的5514端口上去。

2.重启rsyslog

systemctl restart rsyslog

logstash配置文件如下：需要先启动，启动后会开启5514端口，用来侦听。

[root@localhost logstash]# cat rsyslog-logstash.conf
input {
  syslog {
    port => "5514"
  }
}
output {
  stdout {
    codec => rubydebug
  }
}
[root@localhost logstash]#

查看日志

tail -f nohup.out

[2023-08-31T09:51:20,356][INFO ][logstash.inputs.syslog   ][main]
[1ac4f1a43da057380f8444a587ee7cb01fe84a0702afb9d46abc9667eeb0ea0c] Starting syslog tcp listener
{:address=>"0.0.0.0:5514"}
[2023-08-31T09:51:20,390][INFO ][logstash.inputs.syslog   ][main]
[1ac4f1a43da057380f8444a587ee7cb01fe84a0702afb9d46abc9667eeb0ea0c] Starting syslog udp listener
{:address=>"0.0.0.0:5514"}

日志源服务器 10.10.10.56 启动rsyslog客户端

1，vi /etc/rsyslog.conf

*.* @@10.10.10.74:5514 #10.10.10.74本机logstash服务器的IP地址，这个配置时使用rsyslog客户端把本机

的日志信息传输到10.10.10.74服务器的5514端口上去。

2.重启rsyslog

systemctl restart rsyslog

[root@node1 ~]# systemctl restart rsyslog

[root@node1 ~]# service status rsyslog
The service command supports only basic LSB actions (start, stop, restart, try-restart, reload, force-
reload, status). For other actions, please try to use systemctl.
[root@node1 ~]# systemctl status rsyslog
● rsyslog.service - System Logging Service
   Loaded: loaded (/usr/lib/systemd/system/rsyslog.service; enabled; vendor preset: enabled)
   Active: active (running) since 四 2023-08-31 10:00:47 CST; 31s ago

logstash信息输出

{
    "facility_label" => "system",
          "@version" => "1",
         "timestamp" => "Aug 31 10:02:01",
          "facility" => 3,
              "host" => "10.10.10.56",
         "logsource" => "node1",
          "priority" => 30,
        "@timestamp" => 2023-08-31T02:02:01.000Z,
          "severity" => 6,
    "severity_label" => "Informational",
           "message" => "Removed slice User Slice of liuyang.\n",
           "program" => "systemd"
}
{
    "facility_label" => "security/authorization",
          "@version" => "1",
         "timestamp" => "Aug 31 10:02:01",
          "facility" => 10,
              "host" => "10.10.10.56",
         "logsource" => "node1",
          "priority" => 87,
        "@timestamp" => 2023-08-31T02:02:01.000Z,
          "severity" => 7,
    "severity_label" => "Debug",
               "pid" => "17805",
           "message" => "pam_limits(crond:session): unknown limit item 'noproc'\n",
           "program" => "crond"

通过上面的日志输入，发现logstash把接收到的日志进行了详细的划分。会把日志中的时间，主机名，程序，具体信

息拆分成多个字段进行存储。

"timestamp" 为源日志的时间

"@timestamp" 为logstash抓取日志的时间，与上面的时间差了8个小时，这个是时区的配置问题。

**********************************************实验四：读取tcp网络数据

下面的时间配置文件是通过"LogStash::Inputs::TCP"和"LogStash::Filters::Grok"相配合实现实验三rsyslog功能

的日志读取

[root@localhost logstash]# cat tcp-logstash.conf
input {
  tcp {
    port => "5514"
  }
}

filter {
  grok {
    match => {"message" => "%{SYSLOGLINE}"}
  }
}

output {
  stdout {
    codec => rubydebug
  }
}

[root@l

启动logstash服务

nohup /usr/local/logstash/bin/logstash -f /usr/local/logstash/tcp-logstash.conf &

查看日志

2023-08-31T10:08:39,596\]\[INFO \]\[logstash.inputs.tcp \]\[main

e17c63be3a5b12883f975a9f5eaf27f19639714f6267583b2142379ed6c8f22a Starting tcp input listener

{:address=>"0.0.0.0:5514", :ss l_enable=>"false"}

5514端口已启动

客户端同样适用rsyslog，同上一样的配置

logstash 日志查询

{
          "port" => 58526,
       "message" => [
        [0] "<30>Aug 31 10:17:01 node1 systemd: Started Session 351785 of user liuyang.",
        [1] "Started Session 351785 of user liuyang."
    ],
       "program" => "systemd",
     "logsource" => "node1",
          "host" => "10.10.10.56",
     "timestamp" => "Aug 31 10:17:01",
      "@version" => "1",
    "@timestamp" => 2023-08-31T02:10:24.759Z
}
{
          "port" => 58526,
       "message" => [
        [0] "<85>Aug 31 10:17:04 node1 polkitd[1172]: Registered Authentication Agent for unix-
process:20549:1866610077 (system bus name :1.703886 [/usr/bin/pkttyagent --notify-fd 5 --fallback],
object path /org/freedesktop/PolicyKit1/AuthenticationAgent, locale zh_CN.UTF-8)",
        [1] "Registered Authentication Agent for unix-process:20549:1866610077 (system bus name
:1.703886 [/usr/bin/pkttyagent --notify-fd 5 --fallback], object path
/org/freedesktop/PolicyKit1/AuthenticationAgent, locale zh_CN.UTF-8)"
    ],
       "program" => "polkitd",
     "logsource" => "node1",
           "pid" => "1172",
          "host" => "10.10.10.56",
     "timestamp" => "Aug 31 10:17:04",
      "@version" => "1",
    "@timestamp" => 2023-08-31T02:10:26.949Z
}

我们可以看出【0】是完整的信息输出

【1】是经过拆分的，如pid 、logsource、 时间、port都进行拆分出来了

tcp方式和rsyslog类似。

******************************************************

实验五 适用nc 的方式将日志导入到logstash

客户端(日志源)10.10.10.56

服务器（logstash）10.10.10.74服务器logstash配置方法如实验四

[root@localhost logstash]# cat tcp-logstash.conf
input {
  tcp {
    port => "5514"
  }
}

filter {
  grok {
    match => {"message" => "%{SYSLOGLINE}"}
  }
}

output {
  stdout {
    codec => rubydebug
  }
}

客户端命令行窗口输入：

nc 10.10.10.74 5514 </var/log/messages

在logstash上查看日志

{
          "port" => 59322,
       "message" => [
        [0] "Aug 27 12:34:57 node1 supervisord: 2023-08-27 12:34:57,594 INFO supervisord started with
pid 4760",
        [1] "2023-08-27 12:34:57,594 INFO supervisord started with pid 4760"
    ],
       "program" => "supervisord",
     "logsource" => "node1",
          "host" => "10.10.10.56",
     "timestamp" => "Aug 27 12:34:57",
      "@version" => "1",
    "@timestamp" => 2023-08-31T02:16:27.986Z
}
{
          "port" => 59322,
       "message" => [
        [0] "Aug 27 12:34:58 node1 supervisord",
        [1] "supervisord"
    ],
     "logsource" => "node1",
          "host" => "10.10.10.56",
     "timestamp" => "Aug 27 12:34:58",
      "@version" => "1",
    "@timestamp" => 2023-08-31T02:16:27.992Z
}

完成实验。

实验六 编码插件codec

此插件可以放到输入和输出时来处理数据

input -> decode --> filter --> decode --->output decode就是使用codec进行编码

codec支持plain 、json、json_lines等格式。

1.codec插件之plain

plain是一个空解释器，输入什么格式，输出就是什么格式

[root@localhost logstash]# vi codec1-logstash.log
input {
  stdin { 
  }
}

output {
  stdout {
    codec => "plain"
#前面的测试我们都使用rubydebug编码，此编码会以json的格式进行输出
  }
}

root@localhost logstash# vi codec1-logstash.log

root@localhost logstash# /usr/local/logstash/bin/logstash -f /usr/local/logstash/codec1-logstash.log

此处不能用nohup，否则抓取不到

hello             #我用键盘输入的，下面的信息是logstash的输出。
2023-08-31T02:28:31.068Z localhost.localdomain hello
nihao
2023-08-31T02:28:37.572Z localhost.localdomain nihao
仅增加了2个字段一个是时间戳，一个是主机名

和以前使用rubydebug的日志来对比一下

hello #我用键盘输入的，下面的信息是logstash的输出。

{
       "message" => "hello",
      "@version" => "1",
          "host" => "localhost.localdomain",
    "@timestamp" => 2023-08-29T02:22:53.965Z
}

2.codec插件之json

发送给logstash的数据如果是json格式的，那必须在input字段加入codec=> json来解析进来的数据，

如果想让logstash输出为json的格式，可以在output字段加入codec=>json，

[root@localhost logstash]# vi codec2-logstash.log
input {
  stdin { 
  }
}

output {
  stdout {
    codec => "json"   #以json的格式 输出
  }
}

json模式就是key:values格式

3.codec插件之json_lines

若果json文件比较长，需要换行的话，就会使用json_lines编码格式。

实验七 过滤器插件filter

1.grok正则捕获

grok是一个强大的filter插件，通过正则解析任意文本文件，将非结构化的数据弄成结构化的数据，方便查询。

https://help.aliyun.com/zh/sls/user-guide/grok-patterns

GROK的模式参考及示例

grok的语法规则

%{语法:语义}

语法指的就是匹配模式，例如使用number模式可以匹配数字，ip模式会匹配出127.0.0.1样式的IP地址

例如1.输入内容为：172.16.213.132 07/Feb/2018:16:24:19 +0800"GET / HTTP/ 1.1" 403 5039

那么，%{IP:clientip} IP就是语法，要匹配IP地址， clientip为内容

匹配的结果为clientip:172.16.213.132

例如2：

%{HTTPDATE:timestamp}结果为07/feb/2018:16:24 +800

例如3：

%{QS:referrer}匹配的结果

GET / HTTP/ 1.1

以上IP\ Httpdate 、QS都是grok内部定义好的模式，

/usr/local/logstash/vendor/bundle/jruby/2.5.0/gems/logstash-patterns-core-4.1.2/patterns

root@localhost patterns# ls

aws bind exim grok-patterns httpd junos maven mcollective-patterns nagios

rails ruby

bacula bro firewalls haproxy java linux-syslog mcollective mongodb

postgresql redis squid

这个目录下，有很多匹配模式，我们可以直接拿来应用，其中grok-patterns使我们使用的基础匹配模式 vi grok-patterns 显示一小段内容如下： IPV6 ((([0-9A-Fa-f]{1,4}:){7}([0-9A-Fa-f]{1,4}|:))|(([0-9A-Fa-f]{1,4}:){6}(:[0-9A-Fa-f]{1,4}|((25[0- 5]|2[0-4]\d|1\d\d|[1-9]?\d)(\.(25[0-5]|2[0-4]\d|1\d\d|[1-9]?\d)){3})|:))|(([0-9A-Fa-f]{1,4}:){5}(((:[0 -9A-Fa-f]{1,4}){1,2})|:((25[0-5]|2[0-4]\d|1\d\d|[1-9]?\d)(\.(25[0-5]|2[0-4]\d|1\d\d|[1-9]?\d)) {3})|:))|(([0-9A-Fa-f]{1,4}:){4}(((:[0-9A-Fa-f]{1,4}){1,3})|((:[0-9A-Fa-f]{1,4})?:((25[0-5]|2[0-4]\d| 1\d\d|[1-9]?\d)(\.(25[0-5]|2[0-4]\d|1\d\d|[1-9]?\d)){3}))|:))|(([0-9A-Fa-f]{1,4}:){3}(((:[0-9A-Fa-f] {1,4}){1,4})|((:[0-9A-Fa-f]{1,4}){0,2}:((25[0-5]|2[0-4]\d|1\d\d|[1-9]?\d)(\.(25[0-5]|2[0-4]\d|1\d\d|[1 -9]?\d)){3}))|:))|(([0-9A-Fa-f]{1,4}:){2}(((:[0-9A-Fa-f]{1,4}){1,5})|((:[0-9A-Fa-f]{1,4}){0,3}:((25[0- 5]|2[0-4]\d|1\d\d|[1-9]?\d)(\.(25[0-5]|2[0-4]\d|1\d\d|[1-9]?\d)){3}))|:))|(([0-9A-Fa-f]{1,4}:){1}(((: [0-9A-Fa-f]{1,4}){1,6})|((:[0-9A-Fa-f]{1,4}){0,4}:((25[0-5]|2[0-4]\d|1\d\d|[1-9]?\d)(\.(25[0-5]|2[0- 4]\d|1\d\d|[1-9]?\d)){3}))|:))|(:(((:[0-9A-Fa-f]{1,4}){1,7})|((:[0-9A-Fa-f]{1,4}){0,5}:((25[0-5]|2[0- 4]\d|1\d\d|[1-9]?\d)(\.(25[0-5]|2[0-4]\d|1\d\d|[1-9]?\d)){3}))|:)))(%.+)? IPV4 (? ["message","%{IP:clientip}"] } } output { stdout { codec => "rubydebug" } } \[root@localhost logstash\]# /usr/local/logstash/bin/logstash -f /usr/local/logstash/grok1-logstash.log 172.16.213.132 [07/Feb/2018:16:24:19 +0800]"GET / HTTP/ 1.1" 403 5039 #输入 {输出为： "clientip" => "172.16.213.132", "@timestamp" => 2023-08-31T07:36:18.564Z, "@version" => "1", "message" => "172.16.213.132 [07/Feb/2018:16:24:19 +0800]\"GET / HTTP/ 1.1\" 403 5039", "host" => "localhost.localdomain" 测试2： \[root@localhost logstash\]# vi grok2-logstash.log input { stdin { } } filter { grok { match => ["message","%{IP:client-ip} \[%{HTTPDATE:timestamp1}\]%{QS:refer}\ %{NUMBER:nu} %{NUMBER:bytes1}"] } } output { stdout { codec => "rubydebug" } } 输出为： 172.16.213.132 [07/Feb/2018:16:24:19 +0800]"GET / HTTP/ 1.1" 403 5039 { "host" => "localhost.localdomain", "timestamp1" => "07/Feb/2018:16:24:19 +0800", "refer" => "\"GET / HTTP/ 1.1\"", "nu" => "403", "@version" => "1", "message" => "172.16.213.132 [07/Feb/2018:16:24:19 +0800]\"GET / HTTP/ 1.1\" 403 5039", "@timestamp" => 2023-08-31T07:46:51.778Z, "bytes1" => "5039", "client-ip" => "172.16.213.132" 我们已将看到message已经分成5部分了，原有的message可以去掉了，系统中存在连个timestamp，其实@timestamp也 不需要了，这个时间是收集日志的时间。而kibana使用@timestamp这个字段来排序。我们可以将timestamp的值付给 @timestamp [root@localhost logstash]# vi grok-delete-logstash.log input { stdin { } } filter { grok { match => ["message","%{IP:client-ip} \[%{HTTPDATE:timestamp1}\]%{QS:refer}\ %{NUMBER:nu} %{NUMBER:bytes1}"] remove_field => ["message"] } date { match => ["timestamp1","dd/mmm/yyyy:HH:mm:ss Z"] } mutate { remove_field => ["timestamp1"] } } output { stdout { codec => "rubydebug" } } 输出结果 172.16.213.132 [07/Feb/2018:16:24:19 +0800]"GET / HTTP/ 1.1" 403 5039 { "@timestamp" => 2023-08-31T08:16:31.119Z, "refer" => "\"GET / HTTP/ 1.1\"", "nu" => "403", "host" => "localhost.localdomain", "client-ip" => "172.16.213.132", "tags" => [ [0] "_dateparsefailure" ], "bytes1" => "5039", "@version" => "1" 以上使用了grok、date、mutate插件 时间处理模式 DATE date插件 就是将值以什么格式赋值给@timestamp date { match =\> \["timestamp1","dd/mmm/yyyy:HH:mm:ss Z"

}

将timestamp1按照后边dd/mmm/yyyy:HH:mm:ss Z的格式赋值给@timestamp

数据修改插件 mutate

1.正则表达式替换匹配字段

gsub可以通过正则表达式替换字段中匹配到的值，只对字符串段有效，例子

filter {

mutate {

gsub => "filed_name_1","/","_"

#表示将field_name_1属性的字段中所有"/"字符替换成"_"

}

}

实例：root@localhost logstash# cat grok-mutate-logstash.log input {

  stdin {
  }
}

filter {
  grok {
    match => ["message","%{IP:client-ip} \[%{HTTPDATE:timestamp1}\]%{QS:refer}\ %{NUMBER:nu}
%{NUMBER:bytes1}"]
  }

date {
    match => ["timestamp1","dd/mmm/yyyy:HH:mm:ss Z"]
   }

mutate {
    gsub => ["message","/","_"]
  }

}

output {
  stdout {
    codec => "rubydebug"
  }
}

/usr/local/logstash/bin/logstash -f /usr/local/logstash/grok-mutate-logstash.log

172.16.213.132 [07/Feb/2018:16:24:19 +0800]"GET / HTTP/ 1.1" 403 5039
{
         "refer" => "\"GET / HTTP/ 1.1\"",
          "tags" => [
        [0] "_dateparsefailure"
    ],
     "client-ip" => "172.16.213.132",
        "bytes1" => "5039",
       "message" => "172.16.213.132 [07_Feb_2018:16:24:19 +0800]\"GET _ HTTP_ 1.1\" 403 5039",
    "@timestamp" => 2023-08-31T08:35:00.815Z,
      "@version" => "1",
          "host" => "localhost.localdomain",
            "nu" => "403",
    "timestamp1" => "07/Feb/2018:16:24:19 +0800"
}

看到"/"都替换成了"_"

2.分隔字符串为数组

split用分隔符分隔字符串为数组

filter {
  mutate {
    split => ["filed_name_2","|"]
#表示将field_name_1属性的字段中所有"/"字符替换成"_"
  }
}

172.16.213.132|07/Feb/2018:16:24:19 +0800|"GET / HTTP/ 1.1"|403|5039

实例：

[root@localhost logstash]# cat grok-mutate2-logstash.log
input {
  stdin {
  }
}

filter {
  grok {
    match => ["message","%{IP:client-ip} \[%{HTTPDATE:timestamp1}\]%{QS:refer}\ %{NUMBER:nu}
%{NUMBER:bytes1}"]
  }

date {
    match => ["timestamp1","dd/mmm/yyyy:HH:mm:ss Z"]
   }

mutate {
    split => ["message","|"]
  }

}

output {
  stdout {
    codec => "rubydebug"
  }
}

结果：

/usr/local/logstash/bin/logstash -f /usr/local/logstash/grok-mutate2-logstash.log

172.16.213.132|[07/Feb/2018:16:24:19 +0800]|"GET / HTTP/ 1.1"|403|5039
{
       "message" => [
        [0] "172.16.213.132",
        [1] "[07/Feb/2018:16:24:19 +0800]",
        [2] "\"GET / HTTP/ 1.1\"",
        [3] "403",
        [4] "5039"
    ],
    "@timestamp" => 2023-08-31T08:39:17.562Z,
          "host" => "localhost.localdomain",
      "@version" => "1",
          "tags" => [
        [0] "_grokparsefailure"
    ]
}

我们发现message的信息分成了5部分，以后调用以数组的形式调用

3.重命名字段rename

mutate {

rename => {"message","message_new"}

}

}

实例略

4.删除字段remove_field

mutate {

remove_field => "message"

}

}

综合实例：

mutate {

rename => {"nu","number"}

gsub => "refer","/","_"

remove_field => "timestamp1"

split => "client-ip","."

}

重命名 替换 删除 分隔都可以写在一起。

Geoip地址查询归类

geoIP是免费的ip地址归类查询库，可以通过IP地址提供对应的地域信息，包括国别，省市，经纬度等，此插件对可视

化地图和区域统计非常有用。

filter {

geoip {

source => "ip_field"

ip_field字段是输出ip地址的一个字段

}

}

实例：

logstash配置root@localhost logstash# cat grok-geoip.log

input {
  stdin {
  }
}

filter {
  grok {
    match => ["message","%{IP:client-ip} \[%{HTTPDATE:timestamp1}\]%{QS:refer}\ %{NUMBER:nu}
%{NUMBER:bytes1}"]
  }

date {
    match => ["timestamp1","dd/mmm/yyyy:HH:mm:ss Z"]
   }

geoip {
  source => "client-ip"
  }
}

output {
  stdout {
    codec => "rubydebug"
  }
}

172.16.213.132 07/Feb/2018:16:24:19 +0800"GET / HTTP/ 1.1" 403 5039

114.114.114.114 07/Feb/2018:16:24:19 +0800"GET / HTTP/ 1.1" 403 5039

输出结果：

202.97.224.68 [07/Feb/2018:16:24:19 +0800]"GET / HTTP/ 1.1" 403 5039
{
    "timestamp1" => "07/Feb/2018:16:24:19 +0800",
        "bytes1" => "5039",
         "geoip" => {
              "latitude" => 45.75,
           "region_name" => "Heilongjiang",
         "country_code2" => "CN",
          "country_name" => "China",
             "longitude" => 126.65,
              "location" => {
            "lon" => 126.65,
            "lat" => 45.75
        },
         "country_code3" => "CN",
           "region_code" => "HL",
        "continent_code" => "AS",
                    "ip" => "202.97.224.68",
              "timezone" => "Asia/Shanghai"
    },
       "message" => "202.97.224.68 [07/Feb/2018:16:24:19 +0800]\"GET / HTTP/ 1.1\" 403 5039",
    "@timestamp" => 2023-09-01T01:27:01.443Z,
     "client-ip" => "202.97.224.68",
         "refer" => "\"GET / HTTP/ 1.1\"",
            "nu" => "403",
      "@version" => "1",
          "host" => "localhost.localdomain",
          "tags" => [
        [0] "_dateparsefailure"
    ]
}

longitude latitude 经纬度

以上信息有些多，想精简一些

精简geoip信息

geoip {

source => "client-ip"

fields => "ip","country_code3","longitude","latitude","region_name"

#仅将需要保留的域显示出来

}root@localhost logstash# cat grok-geoip2.log

input {
  stdin {
  }
}

filter {
  grok {
    match => ["message","%{IP:client-ip} \[%{HTTPDATE:timestamp1}\]%{QS:refer}\ %{NUMBER:nu}
%{NUMBER:bytes1}"]
  }

date {
    match => ["timestamp1","dd/mmm/yyyy:HH:mm:ss Z"]
   }

geoip {
  source => "client-ip"
  fields => ["ip","country_code3","longitude","latitude","region_name"]
  }
}

output {
  stdout {
    codec => "rubydebug"
  }
}

输出结果：

202.97.224.68 07/Feb/2018:16:24:19 +0800"GET / HTTP/ 1.1" 403 5039

{
            "nu" => "403",
    "@timestamp" => 2023-09-01T01:30:39.227Z,
      "@version" => "1",
          "tags" => [
        [0] "_dateparsefailure"
    ],
         "geoip" => {
                   "ip" => "202.97.224.68",
        "country_code3" => "CN",
             "latitude" => 45.75,
          "region_name" => "Heilongjiang",
            "longitude" => 126.65
    },
       "message" => "202.97.224.68 [07/Feb/2018:16:24:19 +0800]\"GET / HTTP/ 1.1\" 403 5039",
     "client-ip" => "202.97.224.68",
          "host" => "localhost.localdomain",
         "refer" => "\"GET / HTTP/ 1.1\"",
    "timestamp1" => "07/Feb/2018:16:24:19 +0800",
        "bytes1" => "5039"
}

logstash的输出插件output

file 将数据写入磁盘文件

elasticsearch ：把日志数据发送到es集群

graphite：用于存储和绘制数据指标

还支持输出到redis，email，exec，ngios等等

1.标准输出

output {

stdout {

codec => "rubydebug"

}

}

2.保存到文件

output {

file {

path => "/data/log3/%{+yyyy-MM-dd}/%{host}_%{+HH}.log"

}

}

例子：

root@localhost logstash# cat file-log.log

input {
  stdin {
  }
}
output {
  file {
    path => "/data/log3/%{+yyyy-MM-dd}/%{host}_%{+HH}.log"
  }
}

/usr/local/logstash/bin/logstash -f /usr/local/logstash/file-log.log

标准输入信息后

114.114.114.114 07/Feb/2018:16:24:19 +0800\"GET / HTTP/ 1.1\" 403 5039

asdfasdf

/data/log3/2023-09-01/下文件内容：

[root@localhost 2023-09-01]# cat localhost.localdomain_01.log
{"@timestamp":"2023-09-01T01:57:05.295Z","message":"114.114.114.114 [07/Feb/2018:16:24:19 +0800]\"GET /
HTTP/ 1.1\" 403 5039","host":"localhost.localdomain","@version":"1"}
[root@localhost 2023-09-01]# tail -f localhost.localdomain_01.log
{"@timestamp":"2023-09-01T01:57:05.295Z","message":"114.114.114.114 [07/Feb/2018:16:24:19 +0800]\"GET /
HTTP/ 1.1\" 403 5039","host":"localhost.localdomain","@version":"1"}
{"@timestamp":"2023-09-
01T01:57:53.440Z","message":"asdfasdf","host":"localhost.localdomain","@version":"1"}

我们发现输出的内容会在输入的内容上加了一些信息，如@timestamp @version host等属性

如果要让输入和输出一样。我们需要使用codec来格式编码

logstash配置

root@localhost logstash# cat file2-log.log

input {
  stdin {
  }
}
output {
  file {
    path => "/data/log3/%{+yyyy-MM-dd}/%{host}_%{+HH}.log"
    codec => line { format => "%{message}"}
  }
}

root@localhost logstash# /usr/local/logstash/bin/logstash -f /usr/local/logstash/file2-log.log

标准输入

adfasdfasdf

ELK大规模日志实时处理系统从入门到企业应用实战视频课程

查看输出：

root@localhost 2023-09-01# tail -f localhost.localdomain_02.log

adfasdfasdf

ELK大规模日志实时处理系统从入门到企业应用实战视频课程

root@localhost 2023-09-01# pwd

/data/log3/2023-09-01

root@localhost 2023-09-01#

输出与输入一致了。

八 ELK手机apache访问日志的案例

elk收集日子的几种方法

1.不修改源日志的输出格式，而是通过logstash的grok方式进行过滤、清晰，然后输出

优点，对业务系统无影响，缺点是logstash可能会有瓶颈。

2.修改源日志的输出格式，按要求的格式改变源日志格式进行输出，logstash仅收集和传输。

优点：减轻了logstash的压力，但是需要一定的工作量去处理源日志格式。

elk收集apache日志应用架构

apache（filebeat） -- kafka（zookeeper） -- logstash -- ES集群

使用第二种方式，用改变源日志输出格式来处理。