搭建部署分布式ELK平台

Posted 2020-10-22

tags:

篇首语：本文由小常识网(cha138.com)小编为大家整理，主要介绍了搭建部署分布式ELK平台相关的知识，希望对你有一定的参考价值。

logstash

? logstash 是什么

– logstash是一个数据采集、加工处理以及传输的工具

? logstash 特点:

– 所有类型的数据集中处理

– 不同模式和格式数据的正常化

– 自定义日志格式的迅速扩展

– 为自定义数据源轻松添加插件

? logstash 安装

– Logstash 依赖 java 环境,需要安装 java-1.8.0-openjdk

– Logstash 没有默认的配置文件,需要手动配置

– logstash 安装在 /opt/logstash 目录下

# rpm -ivh logstash-2.3.4-1.noarch.rpm

# rpm -qc logstash

/etc/init.d/logstash

/etc/logrotate.d/logstash

/etc/sysconfig/logstash

// 查看logstash 模块插件列表

# /opt/logstash/bin/logstash-plugin list

logstash-codec-collectd

logstash-codec-json

.. ..

logstash-filter-anonymize

logstash-filter-checksum

.. ..

logstash-input-beats

logstash-input-exec

.. ..

logstash-output-cloudwatch

logstash-output-csv

.. ..

logstash-patterns-core

第一列表示是 logstash 的模块

第二列表示在那个区域段执行 codec 属于编解码是字符编码类型的在全部的区域段可以运行

? logstash 工作结构

– { 数据源 } ==>

– input { } ==> //收集日志

– filter { } ==> //日志处理整理格式

– output { } ==> //日志输出

– { ES }

? logstash 里面的类型

– 布尔值类型: ssl_enable => true

– 字节类型:bytes => "1MiB"

– 字符串类型: name => "xkops"

– 数值类型: port => 22

– 数组: match => ["datetime","UNIX"]

– 哈希: options => {k => "v",k2 => "v2"}

– 编码解码: codec => "json"

– 路径: file_path => "/tmp/filename"

– 注释: #

? logstash 条件判断

– 等于: ==

– 不等于: !=

– 小于: <

– 大于: >

– 小于等于: <=

– 大于等于: >=

– 匹配正则: =~

– 不匹配正则: !~

logstash 条件判断

– 包含: in

– 不包含: not in

– 与: and

– 或: or

– 非与: nand

– 非或: xor

– 复合表达式: ()

– 取反符合: !()

配置 logastash

# cd /etc/logstash/

// logstash 默认没有配置文件

# vim logstash.conf

input{

stdin{} //标准输入

}

filter{}

output{

stdout{} // 标准输出

}

# /opt/logstash/bin/logstash -f logstash.conf // 相当cat

Settings: Default pipeline workers: 1

Pipeline main started

Hello word!!!

2018-01-26T13:21:22.031Z 0.0.0.0 Hello word!!!

– 上页的配置文件使用了 logstash-input-stdin 和logstash-output-stdout 两个插件

查看插件具体使用方法的方式是 https://github.com/logstash-plugins

技术分享图片

练习1

# vim logstash.conf

input{

stdin{ codec => "json" }

}

filter{}

output{

stdout{}

}

# /opt/logstash/bin/logstash -f logstash.conf

Settings: Default pipeline workers: 1

Pipeline main started

12345 // 123456 不是 json 模式所有报错

A plugin had an unrecoverable error. Will restart this plugin.

Plugin: <LogStash::Inputs::Stdin codec=><LogStash::Codecs::JSON charset=>"UTF-8">>

Error: can't convert String into Integer {:level=>:error}

'abc'

2018-01-26T13:43:17.840Z 0.0.0.0 'abc'

'{"a":1,"b":2}'

2018-01-26T13:43:46.889Z 0.0.0.0 '{"a":1,"b":2}'

练习2

# vim logstash.conf

input{

stdin{ codec => "json" }

}

filter{}

output{

stdout{ codec => "rubydebug"} //调试数据的模式

}

# /opt/logstash/bin/logstash -f logstash.conf

Settings: Default pipeline workers: 1

Pipeline main started

'"aaa"'

{

"message" => "'\"aaa\"'",

"tags" => [

[0] "_jsonparsefailure"

"@version" => "1",

"@timestamp" => "2018-01-26T13:52:45.307Z",

"host" => "0.0.0.0"

}

{"aa":1,"bb":2}

{

"aa" => 1,

"bb" => 2,

"@version" => "1",

"@timestamp" => "2018-01-26T13:53:00.452Z",

"host" => "0.0.0.0"

}

练习3

# touch /tmp/a.log /tmp/b.log

# vim logstash.conf

input{

file { //监控文件

path => ["/tmp/a.log","/tmp/b.log"] //监控文件路径

type => "testlog" //声明文件类型

}

filter{}

output{

stdout{ codec => "rubydebug"}

}

# /opt/logstash/bin/logstash -f logstash.conf

Settings: Default pipeline workers: 1

Pipeline main started

... //开始监控日志文件

//切换到另一个终端测试

# cd /tmp/

// 为日志文件添加内容

# echo 12345 > a.log

// 这时在会看见终端输出监控信息

{

"message" => "12345",

"@version" => "1",

"@timestamp" => "2018-01-27T00:45:15.470Z",

"path" => "/tmp/a.log",

"host" => "0.0.0.0",

"type" => "testlog"

}

# echo b123456 > b.log

{

"message" => "b123456",

"@version" => "1",

"@timestamp" => "2018-01-27T00:45:30.487Z",

"path" => "/tmp/b.log",

"host" => "0.0.0.0",

"type" => "testlog"

}

# echo c123456 >> b.log

{

"message" => "c123456",

"@version" => "1",

"@timestamp" => "2018-01-27T00:45:45.501Z",

"path" => "/tmp/b.log",

"host" => "0.0.0.0",

"type" => "testlog"

}

// 默认记录读取位置的文件在管理员家目录 .sincedb_ 后面是一串哈希数

# cat /root/.sincedb_ab3977c541d1144f701eedeb3af4956a

3190503 0 64768 6

3190504 0 64768 16

# du -b /tmp/a.log du -b 查看文件类型

6/tmp/a.log

# du -b /tmp/b.log

16/tmp/b.log

// 进行优化

# vim logstash.conf

input{

file {

start_position => "beginning" //设置当记录位置的库文件不存在时从文件开始读取

sincedb_path => "/var/lib/logstash/sincedb-access" //记录位置的库文件默认放在每个用户下的所以将记录位置的库文件固定位置

path => ["/tmp/a.log","/tmp/b.log"]

type => "testlog"

}

filter{}

output{

stdout{ codec => "rubydebug"}

}

# rm -rf /root/.sincedb_ab3977c541d1144f701eedeb3af4956a

# /opt/logstash/bin/logstash -f logstash.conf

Settings: Default pipeline workers: 1

Pipeline main started

{

"message" => "12345",

"@version" => "1",

"@timestamp" => "2018-01-27T10:44:54.890Z",

"path" => "/tmp/a.log",

"host" => "0.0.0.0",

"type" => "testlog"

}

{

"message" => "b123456",

"@version" => "1",

"@timestamp" => "2018-01-27T10:44:55.242Z",

"path" => "/tmp/b.log",

"host" => "0.0.0.0",

"type" => "testlog"

}

{

"message" => "c123456",

"@version" => "1",

"@timestamp" => "2018-01-27T10:44:55.242Z",

"path" => "/tmp/b.log",

"host" => "0.0.0.0",

"type" => "testlog"

}

练习 tcp udp 插件

# vim logstash.conf

input{

tcp {

host => "0.0.0.0"

port => 8888

type => "tcplog"

}

udp {

host => "0.0.0.0"

port => 9999

type => "udplog"

}

filter{}

output{

stdout{ codec => "rubydebug"}

}

# /opt/logstash/bin/logstash -f logstash.conf

Settings: Default pipeline workers: 1

Pipeline main started

...

//在另一个终端查看监听端口

# netstat -pantu | grep -E "(8888|9999)"

tcp6 0 0 :::8888 :::* LISTEN 3098/java

udp6 0 0 :::9999 :::* 3098/java

模拟客户端发送数据报文

// 发送tcp 数据报文 exec 改变当前文件描述符标准输入标准输出重新设置

# exec 9<>/dev/tcp/192.168.4.10/8888 //打开建立连接

# echo "hello world" >&9 //发送 hello world 字符串给连接

# exec 9<&- // 关闭连接

// 接收到连接

{

"message" => "hello world",

"@version" => "1",

"@timestamp" => "2018-01-27T11:01:35.356Z",

"host" => "192.168.4.11",

"port" => 48654,

"type" => "tcplog"

}

// 发送udp 数据报文

# exec 7<>/dev/udp/192.168.4.10/9999

# echo "is udp log" >&7

# exec 7<&-

// 接收到连接

{

"message" => "is udp log\n",

"@version" => "1",

"@timestamp" => "2018-01-27T11:05:18.850Z",

"type" => "udplog",

"host" => "192.168.4.11"

}

// 发送文件

# exec 8<>/dev/udp/192.168.4.10/9999

# cat /etc/hosts >&8

# exec 9<&-

// 接受到信息

{

"message" => "127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4\n::1 localhost localhost.localdomain localhost6 localhost6.localdomain6\n192.168.4.11\tes1\n192.168.4.12\tes2\n192.168.4.13\tes3\n192.168.4.14\tes4\n192.168.4.15\tes5\n",

"@version" => "1",

"@timestamp" => "2018-01-27T11:10:31.099Z",

"type" => "udplog",

"host" => "192.168.4.11"

}

syslog 插件练习

# vim logstash.conf

input{

syslog{

host => "192.168.4.10"

port => 514 //系统日志默认端口

type => "syslog"

}

filter{}

output{

stdout{ codec => "rubydebug"}

}

# /opt/logstash/bin/logstash -f logstash.conf

Settings: Default pipeline workers: 1

Pipeline main started

. . .

# netstat -pantu | grep 514

tcp6 0 0 192.168.4.10:514 :::* LISTEN 3545/java

udp6 0 0 192.168.4.10:514 :::* 3545/java

//在客户端主机上自定义日志文件发给logstash 主机

# vim /etc/rsyslog.conf

# sed -n '74p' /etc/rsyslog.conf

local0.info@@192.168.4.10:514

# systemctl restart rsyslog.service

# logger -p local0.info -t testlog "hello world" // 发送一条测试日志

// logstash 主机收到日志

{

"message" => "hello world\n",

"@version" => "1",

"@timestamp" => "2018-01-27T11:29:03.000Z",

"type" => "syslog",

"host" => "192.168.4.11",

"priority" => 134,

"timestamp" => "Jan 27 06:29:03",

"logsource" => "es1",

"program" => "testlog",

"severity" => 6,

"facility" => 16,

"facility_label" => "local0",

"severity_label" => "Informational"

}

扩展如果想要把登录日志发给 logstash

# vim /etc/rsyslog.conf

# sed -n '75p' /etc/rsyslog.conf

authpriv.*@@192.168.4.10:514

# systemctl restart rsyslog.service

//测试登录登出

# exit

登出

Connection to 192.168.4.11 closed.

# ssh -X [email protected]

[email protected]'s password:

Last login: Sat Jan 27 05:27:21 2018 from 192.168.4.254

//实时接受的日志

{

"message" => "Accepted password for root from 192.168.4.254 port 50820 ssh2\n",

"@version" => "1",

"@timestamp" => "2018-01-27T11:32:07.000Z",

"type" => "syslog",

"host" => "192.168.4.11",

"priority" => 86,

"timestamp" => "Jan 27 06:32:07",

"logsource" => "es1",

"program" => "sshd",

"pid" => "3734",

"severity" => 6,

"facility" => 10,

"facility_label" => "security/authorization",

"severity_label" => "Informational"

}

{

"message" => "pam_unix(sshd:session): session opened for user root by (uid=0)\n",

"@version" => "1",

"@timestamp" => "2018-01-27T11:32:07.000Z",

"type" => "syslog",

"host" => "192.168.4.11",

"priority" => 86,

"timestamp" => "Jan 27 06:32:07",

"logsource" => "es1",

"program" => "sshd",

"pid" => "3734",

"severity" => 6,

"facility" => 10,

"facility_label" => "security/authorization",

"severity_label" => "Informational"

}

? filter grok插件

– 解析各种非结构化的日志数据插件

– grok 使用正则表达式把飞结构化的数据结构化

– 在分组匹配,正则表达式需要根据具体数据结构编写

– 虽然编写困难,但适用性极广

– 几乎可以应用于各类数据

? grok 正则分组匹配

– 匹配 ip 时间戳 和请求方法

"(?<ip>(\d+\.){3}\d+) \S+ \S+

(?<time>.*\])\s+\"(?<method>[A-Z]+)"]

– 使用正则宏

%{IPORHOST:clientip} %{HTTPDUSER:ident} %{USER:auth}

\[%{HTTPDATE:timestamp}\] \"%{WORD:verb}

– 最终版本

%{COMMONAPACHELOG} \"(?<referer>[^\"]+)\"

\"(?<UA>[^\"]+)\"

练习匹配Apache 日志

# cd /opt/logstash/vendor/bundle/jruby/1.9/gems/logstash-patterns-core-2.0.5/patterns/

# vim grok-patterns //日志宏定义仓库

....

COMMONAPACHELOG %{IPORHOST:clientip} %{HTTPDUSER:ident} %{USER:auth} \[%{HTTPDATE:timestamp}\] "(?:%{WORD:verb} %{NOTSPACE:request}(?: HTTP/%{NUMBER:httpversion})?|%{DATA:rawrequest})" %{NUMBER:response} (?:%{NUMBER:bytes}|-)

....

# vim /tmp/test.log //在测试文件中添加一条日志

220.181.108.115 - - [11/Jul/2017:03:07:16 +0800] "GET /%B8%DF%BC%B6%D7%E2%C1%DE%D6%F7%C8%CE/QYQiu_j.html HTTP/1.1" 200 20756 "-" "Mozilla/5.0 (compatible; Baiduspider/2.0; +http://www.baidu.com/search/spider.html)" "-" zxzs.buildhr.com 558742

# vim logstash.conf

input{

file{

start_position => "beginning"

sincedb_path => "/dev/null" //为了调试方便

path => [ "/tmp/test.log" ]

type => 'filelog'

}

filter{

grok{

match => ["message","%{COMMONAPACHELOG}"]

}

output{

stdout{ codec => "rubydebug"}

}

# /opt/logstash/bin/logstash -f logstash.conf Settings: Default pipeline workers: 1

Pipeline main started

{

"message" => "220.181.108.115 - - [11/Jul/2017:03:07:16 +0800] \"GET /%B8%DF%BC%B6%D7%E2%C1%DE%D6%F7%C8%CE/QYQiu_j.html HTTP/1.1\" 200 20756 \"-\" \"Mozilla/5.0 (compatible; Baiduspider/2.0; +http://www.baidu.com/search/spider.html)\" \"-\" zxzs.buildhr.com 558742",

"@version" => "1",

"@timestamp" => "2018-01-27T12:03:10.363Z",

"path" => "/tmp/test.log",

"host" => "0.0.0.0",

"type" => "filelog",

"clientip" => "220.181.108.115",

"ident" => "-",

"auth" => "-",

"timestamp" => "11/Jul/2017:03:07:16 +0800",

"verb" => "GET",

"request" => "/%B8%DF%BC%B6%D7%E2%C1%DE%D6%F7%C8%CE/QYQiu_j.html",

"httpversion" => "1.1",

"response" => "200",

"bytes" => "20756"

}

{

"message" => "",

"@version" => "1",

"@timestamp" => "2018-01-27T12:03:10.584Z",

"path" => "/tmp/test.log",

"host" => "0.0.0.0",

"type" => "filelog",

"tags" => [

[0] "_grokparsefailure"

]

}

编写分析日志正则时先区宏库中寻找找不到可以去百度尽量不要自己手动写会很累

练习同时解析不同日志

# vim logstash.conf

input{

file{

start_position => "beginning"

sincedb_path => "/dev/null"

path => [ "/tmp/test.log" ]

type => 'filelog'

}

file{

start_position => "beginning"

sincedb_path => "/dev/null"

path => [ "/tmp/test.json" ]

type => 'jsonlog'

codec => 'json'

}

filter{

if [type] == "filelog"{

grok{

match => ["message","%{COMMONAPACHELOG}"]

}

output{

stdout{ codec => "rubydebug"}

}

[[email protected] logstash]# /opt/logstash/bin/logstash -f logstash.conf

Settings: Default pipeline workers: 1

Pipeline main started

{

"@version" => "1",

"@timestamp" => "2018-01-27T12:22:06.481Z",

"path" => "/tmp/test.log",

"host" => "0.0.0.0",

"type" => "filelog",

"clientip" => "220.181.108.115",

"ident" => "-",

"auth" => "-",

"timestamp" => "11/Jul/2017:03:07:16 +0800",

"verb" => "GET",

"request" => "/%B8%DF%BC%B6%D7%E2%C1%DE%D6%F7%C8%CE/QYQiu_j.html",

"httpversion" => "1.1",

"response" => "200",

"bytes" => "20756"

}

{

"message" => "",

"@version" => "1",

"@timestamp" => "2018-01-27T12:22:06.834Z",

"path" => "/tmp/test.log",

"host" => "0.0.0.0",

"type" => "filelog",

"tags" => [

[0] "_grokparsefailure"

]

}

{

"@timestamp" => "2015-05-18T12:28:25.013Z",

"ip" => "79.1.14.87",

"extension" => "gif",

"response" => "200",

"geo" => {

"coordinates" => {

"lat" => 35.16531472,

"lon" => -107.9006142

"src" => "GN",

"dest" => "US",

"srcdest" => "GN:US"

"@tags" => [

[0] "success",

[1] "info"

"utc_time" => "2015-05-18T12:28:25.013Z",

"referer" => "http://www.slate.com/warning/b-alvin-drew",

"agent" => "Mozilla/5.0 (X11; Linux i686) AppleWebKit/534.24 (KHTML, like Gecko) Chrome/11.0.696.50 Safari/534.24",

"clientip" => "79.1.14.87",

"bytes" => 774,

"host" => "motion-media.theacademyofperformingartsandscience.org",

"request" => "/canhaz/james-mcdivitt.gif",

"url" => "https://motion-media.theacademyofperformingartsandscience.org/canhaz/james-mcdivitt.gif",

"@message" => "79.1.14.87 - - [2015-05-18T12:28:25.013Z] \"GET /canhaz/james-mcdivitt.gif HTTP/1.1\" 200 774 \"-\" \"Mozilla/5.0 (X11; Linux i686) AppleWebKit/534.24 (KHTML, like Gecko) Chrome/11.0.696.50 Safari/534.24\"",

"spaces" => "this is a thing with lots of spaces wwwwoooooo",

"xss" => "<script>console.log(\"xss\")</script>",

"headings" => [

[0] "<h3>charles-bolden</h5>",

[1] "http://www.slate.com/success/barry-wilmore"

"links" => [

[0] "[email protected]",

[1] "http://facebook.com/info/anatoly-solovyev",

[2] "www.www.slate.com"

"relatedContent" => [],

"machine" => {

"os" => "osx",

"ram" => 8589934592

"@version" => "1",

"path" => "/tmp/test.json",

"type" => "jsonlog"

}

真正在项目中可以和开放人员商量让其提供json 格式的日志这样能轻松很多

output ES 插件

调试成功后,把数据写入 ES 集群

# cat logstash.conf

input{

file{

start_position => "beginning"

sincedb_path => "/dev/null"

path => [ "/tmp/test.log" ]

type => 'filelog'

}

file{

start_position => "beginning"

sincedb_path => "/dev/null"

path => [ "/tmp/test.json" ]

type => 'jsonlog'

codec => 'json'

}

filter{

if [type] == "filelog"{

grok{

match => ["message","%{COMMONAPACHELOG}"]

}

output{

if [type] == "filelog"{

elasticsearch {

hosts => ["192.168.4.11:9200","192.168.4.12:9200"]

index => "weblog"

flush_size => 2000

idle_flush_time => 10

}

# /opt/logstash/bin/logstash -f logstash.conf

Settings: Default pipeline workers: 1

Pipeline main started

技术分享图片

以上是关于搭建部署分布式ELK平台的主要内容，如果未能解决你的问题，请参考以下文章

搭建部署分布式ELK平台

ELK+Filebeat+Kafka分布式日志管理平台搭建

基于Kafka+ELK搭建海量日志平台

Centos7下ELK+Redis日志分析平台的集群环境部署记录

ELK搭建实时日志分析平台

ELK日志分析平台搭建----ELASTICSEARCH

搭建部署 分布式ELK平台

搭建部署分布式ELK平台