ELK Stack
ELK 介绍
LOG有多重要这个不言而喻, 面对如此大量的数据,又是分布在不同地方,如何快速准确的查找日志?使用传统的方法,去登陆到一台台机器上查看?这种方法无疑显得非常笨拙和低效了。于是一些聪明人就提出了建立一套集中式的方法,把不同来源的数据集中整合到一个地方。
一个完整的集中式日志系统,是离不开以下几个主要特点的
收集-能够采集多种来源的日志数据
传输-能够稳定的把日志数据传输到中央系统
存储-如何存储日志数据
分析-可以支持 UI 分析
警告-能够提供错误报告,监控机制
基于上述思路,于是许多产品或方案就应运而生了。比如,简单的 Rsyslog ,Syslog-ng ;商业化的 Splunk ;开源的有 FaceBook 公司的 Scribe,Apache 的 Chukwa ,Linkedin 的 Kafak ,Cloudera 的 Fluentd ,ELK 等等。
在上述产品中,Splunk 是一款非常优秀的产品,但是它是商业产品,价格昂贵,让许多人望而却步。
直到 ELK 的出现,让大家又多了一种选择。相对于其他几款开源软件来说,本文重点介绍 ELK 。
ELK 不是一款软件,而是 Elasticsearch 、Logstash 和 Kibana 三种软件产品的首字母缩写。这三者都是开源软件,通常配合使用,而且又先后归于 Elastic.co 公司名下,所以被简称为 ELK Stack
。
架构
Logstash 读取Log
发送至 Elasticsearch , kibana 通过 Elasticsearch 提供的RestfulAPI
查询日志。
可以当作一个MVC
模型,Logstash 是 Controller
层,Elasticsearch]2 是一个 Model
层,kibana 是 View
层。
Elasticsearch
安装
# 不能使用ROOT用户启动,所以创建一个新用户
[root@WEB-PM0121 ~] groupadd elk # 添加用户组
[root@WEB-PM0121 ~] useradd -g elk elk # 添加用户到指定用户组
[root@WEB-PM0121 ~] passwd elk # 为指定用户设置密码
[root@WEB-PM0121 bin]# su elk # 切换用户
[elk@WEB-PM0121 ~]# java -version # 查看JAVA版本
openjdk version "1.8.0_151"
OpenJDK Runtime Environment (build 1.8 .0 _151-b12)
OpenJDK 64 -Bit Server VM (build 25.151 -b12, mixed mode)
sudo yum install java-1.8 .0 -openjdk #如果没有则需要安装
# 下载Elasticsearch
[elk@WEB-PM0121 ~]# wget https://artifacts.elastic.co/downloads/elasticsearch/elasticsearch-6.2.4.tar.gz
--2018 -05 -16 14 :45 :50 -- https://artifacts.elastic.co/downloads/elasticsearch/elasticsearch-6.2 .4 .tar.gz
Resolving artifacts.elastic.co... 54.235 .171 .120 , 107.21 .237 .95 , 107.21 .253 .15 , ...
Connecting to artifacts.elastic.co|54.235 .171 .120 |:443. .. connected.
HTTP request sent, awaiting response... 200 OK
Length: 29056810 (28 M) [binary /octet-stream]
Saving to : “elasticsearch-6.2 .4 .tar.gz.2 ”
72 % [=========================================================> ] 21 ,151 ,222 1.22 M/s eta 9 s
tar xzvf elasticsearch-6.2 .4 .tar.gz # 解压
# 目录结构
[elk@WEB-PM0121 ~]# cd elasticsearch-6.2.4
[elk@WEB-PM0121 elasticsearch-6.2 .4 ]# pwd
/home/chenxu/elasticsearch-6.2 .4
[elk@WEB-PM0121 elasticsearch-6.2 .4 ]# ls
bin config data lib LICENSE.txt logs modules NOTICE.txt plugins README.textile vi
[elk@WEB-PM0121 elasticsearch-6.2 .4 ]#
# 修改配置文件
[elk@WEB-PM0121 elasticsearch-6.2 .4 ]# cd config
[elk@WEB-PM0121 config]# vi elasticsearch.yml
cluster.name: cxelk # 友好名称
network.host: 0.0 .0 .0 # 要不然只能本机访问
# 启动
[elk@WEB-PM0121 config]# cd ../bin
[elk@WEB-PM0121 bin]# ./elasticsearch
# 默认是前台启动,可以用./elasticsearch& 或者 ./elasticsearch -d 后端启动
# 验证访问,出现出现JSON则证明启动成功
[root@WEB-PM0121 bin]# curl 'http://10.12.54.127:9200'
{
"name" : "SvJ09aS" ,
"cluster_name" : "cxelk" ,
"cluster_uuid" : "WbsI8yKWTsKUwhU8Os8vJQ" ,
"version" : {
"number" : "6.2.4" ,
"build_hash" : "ccec39f" ,
"build_date" : "2018-04-12T20:37:28.497551Z" ,
"build_snapshot" : false ,
"lucene_version" : "7.2.1" ,
"minimum_wire_compatibility_version" : "5.6.0" ,
"minimum_index_compatibility_version" : "5.0.0"
},
"tagline" : "You Know, for Search"
}
常见问题
ERROR: bootstrap checks failed:max file descriptors [4096] for elasticsearch process likely too low, increase to at least [65536]
原因:无法创建本地文件问题,用户最大可创建文件数太小
解决方案:
切换到root用户,编辑limits.conf配置文件, 添加类似如下内容:
vi /etc/security/limits.conf
添加如下内容:
* soft nofile 65536
* hard nofile 131072
* soft nproc 2048
* hard nproc 4096
max number of threads [1024] for user [es] likely too low, increase to at least [2048]
原因:无法创建本地线程问题,用户最大可创建线程数太小
解决方案:切换到root用户,进入limits.d目录下,修改90-nproc.conf 配置文件。
vi /etc/security/limits.d/90-nproc.conf
修改 * soft nproc 1024 为 * soft nproc 2048
max virtual memory areas vm.max_map_count [65530] likely too low, increase to at least [262144]
原因:最大虚拟内存太小
解决方案:切换到root用户下,修改配置文件sysctl.conf
vi /etc/sysctl.conf
添加下面配置:vm.max_map_count=655360
并执行命令:sysctl -p
Exception in thread “main” 2017-11-10 06:29:49,106 main ERROR No log4j2 configuration file found. Using default configuration: logging only errors to the console. Set system property ‘log4j2.debug’ to show Log4j2 internal initialization logging.ElasticsearchParseException[malformed, expected settings to start with ‘object’, instead was [VALUE_STRING]]
原因:elasticsearch.yml中的配置项的格式有问题
解决方案:请尽量保持冒号前面没空格,后面一个空格,不要用tab键
bootstrap.memory_lock: false
关闭 Elasticsearch
[root@WEB-PM0121 bin]
[root@WEB-PM0121 bin]
Elasticsearch-head
[elk@WEB-PM0121]
[elk@WEB-PM0121]
[elk@WEB-PM0121]
[elk@WEB-PM0121 elasticsearch-head]
[elk@WEB-PM0121 elasticsearch-head]
[root@WEB-PM0121 elasticsearch-head]
Kibana
[root@WEB-PM0121 ~]# wget https://artifacts.elastic.co/downloads/kibana/kibana-6.2.4-linux-x86_64.tar.gz # 下载Kibana
[root@WEB-PM0121 ~]# tar xzvf kibana-6.2.4-linux-x # 解压
# 目录结构
[elk@WEB-PM0121 kibana-6.2 .4 -linux-x86_64]$ cd ..
[elk@WEB-PM0121 chenxu]$ cd kibana-6.2 .4 -linux-x86_64
[elk@WEB-PM0121 kibana-6.2 .4 -linux-x86_64]$ ll
total 1196
drwxr-xr-x 2 1000 1000 4096 Apr 13 04 :57 bin
drwxrwxr-x 2 1000 1000 4096 May 14 15 :18 config
drwxrwxr-x 2 1000 1000 4096 May 14 15 :07 data
-rw-rw-r-- 1 1000 1000 562 Apr 13 04 :57 LICENSE.txt
drwxrwxr-x 6 1000 1000 4096 Apr 13 04 :57 node
drwxrwxr-x 909 1000 1000 36864 Apr 13 04 :57 node_modules
-rw-rw-r-- 1 1000 1000 1134238 Apr 13 04 :57 NOTICE.txt
drwxrwxr-x 3 1000 1000 4096 Apr 13 04 :57 optimize
-rw-rw-r-- 1 1000 1000 721 Apr 13 04 :57 package.json
drwxrwxr-x 2 1000 1000 4096 Apr 13 04 :57 plugins
-rw-rw-r-- 1 1000 1000 4772 Apr 13 04 :57 README.txt
drwxr-xr-x 15 1000 1000 4096 Apr 13 04 :57 src
drwxrwxr-x 5 1000 1000 4096 Apr 13 04 :57 ui_framework
drwxr-xr-x 2 1000 1000 4096 Apr 13 04 :57 webpackShims
# 修改配置文件
[elk@WEB-PM0121 kibana-6.2 .4 -linux-x86_64]$ cd config
[elk@WEB-PM0121 config]$ vi kibana.yml
# 修改以下配置节点
server.port: 5601
server.host: "0.0.0.0"
elasticsearch.url: "http://localhost:9200" # elasticsearch 端口
kibana.index: ".kibana"
# 启动Kibana
[elk@WEB-PM0121 config]$ cd ../bin
[elk@WEB-PM0121 config]$ ./kibana
# 验证Kibana
[elk@WEB-PM0121 bin]$ curl
<script>var hashRoute =
var defaultRoute =
var hash = window.location.hash;
if (hash.length) {
window.location = hashRoute + hash;
} else {
window.location = defaultRoute;
Logstash
由于生产系统基于.NET
,所以 Logstash
在 Windows
下部署, 在 Logstash
下载页面下载对应的压缩包 。
配置文件格式
Logstash
需要一个配置管理输入、过滤器和输出相关的配置。配置内容格式如下
# 输入
input {
...
}
# 过滤器
filter {
...
}
# 输出
output {
...
}
测试输入输出
测试一下输入输出, 在Logstash
中的config
文件夹下新建 logstash_test.conf
键入测试代码
input { stdin { } } output { stdout {} }
E:\Dev\ELK\logstash-6.2 .3 \bin>logstash -f ../config/logstash_test.conf # 启动并指定配置文件
Sending Logstash
d via log4j2.properties
[2018 -05 -17 T14:04 :26 ,229 ][INFO ][logstash.modules.scaffold] Initializing module
{:module_name=>"fb_apache" , :directory=>"E:/Dev/ELK/logstash-6.2.3/modules/fb_ap
ache/configuration"}
[2018 -05 -17 T14:04 :26 ,249 ][INFO ][logstash.modules.scaffold] Initializing module
{:module_name=>"netflow" , :directory=>"E:/Dev/ELK/logstash-6.2.3/modules/netflow
/configuration"}
[2018 -05 -17 T14:04 :26 ,451 ][WARN ][logstash.config.source.multilocal] Ignoring the
[2018 -05 -17 T14:04 :27 ,193 ][INFO ][logstash.runner ] Starting Logstash {"
logstash.version"=>"6.2.3" }
[2018 -05 -17 T14:04 :28 ,016 ][INFO ][logstash.agent ] Successfully started
Logstash API endpoint {:port=>9600 }
[2018 -05 -17 T14:04 :29 ,038 ][INFO ][logstash.pipeline ] Starting pipeline {:
pipeline_id=>"main" , "pipeline.workers" =>4 , "pipeline.batch.size" =>125 , "pipelin
e.batch.delay"=>50 }
[2018 -05 -17 T14:04 :29 ,164 ][INFO ][logstash.pipeline ] Pipeline started suc
cesfully {:pipeline_id=>"main" , :thread=>"#<Thread:0x47319180 run>" }
The stdin plugin is now waiting for input:
[2018 -05 -17 T14:04 :29 ,378 ][INFO ][logstash.agent ] Pipelines running {:
count=>1 , :pipelines=>["main" ]}
123 # 输入测试数据
2018 -05 -17 T06:05 :00.467 Z PC201801151216 123 # 输出的结果
456 # 输入测试数据
2018 -05 -17 T06:05 :04.877 Z PC201801151216 456 # 输出的结果
发送至Elasticsearch
我们需要从文件中读取并发送到 elasticsearch
中。
在Logstash
中的config
文件夹下新建logstash.conf
键入代码
input {
file { # 指定文件模式
path => "E:/WebSystemLog/*" # 测试日志文件
start_position => "beginning"
}
}
output {
elasticsearch{
hosts=> ["http://10.12.54.127:9200" ]
index => "chenxu-%{+YYYY.MM.dd}"
}
stdout {} # 控制台打印
}
在Logstash
根目录新建一个run.bat
方便我们启动Logstash
键入代码
./bin/logstash.bat -f ./config/logstash.conf
E:\Dev\ELK\logstash-6.2 .3 \bin>cd ..
E:\Dev\ELK\logstash-6.2 .3 >run
E:\Dev\ELK\logstash-6.2 .3 >./bin/logstash.bat -f ./config/logstash.conf
Sending Logstash's logs to E:/Dev/ELK/logstash-6.2.3/logs which is now configure
d via log4j2.properties
[2018-05-17T15:17:36,317][INFO ][logstash.modules.scaffold] Initializing module
{:module_name=>"fb_apache", :directory=>"E:/Dev/ELK/logstash-6.2.3/modules/fb_ap
ache/configuration"}
[2018-05-17T15:17:36,334][INFO ][logstash.modules.scaffold] Initializing module
{:module_name=>"netflow", :directory=>"E:/Dev/ELK/logstash-6.2.3/modules/netflow
/configuration"}
[2018-05-17T15:17:36,533][WARN ][logstash.config.source.multilocal] Ignoring the
'pipelines.yml' file because modules or command line options are specified
[2018-05-17T15:17:37,127][INFO ][logstash.runner ] Starting Logstash {"
logstash.version"=>"6.2.3"}
[2018-05-17T15:17:37,682][INFO ][logstash.agent ] Successfully started
Logstash API endpoint {:port=>9600}
[2018-05-17T15:17:39,774][INFO ][logstash.pipeline ] Starting pipeline {:
pipeline_id=>"main", "pipeline.workers"=>4, "pipeline.batch.size"=>125, "pipelin
e.batch.delay"=>50}
[2018-05-17T15:17:40,170][INFO ][logstash.outputs.elasticsearch] Elasticsearch p
ool URLs updated {:changes=>{:removed=>[], :added=>[http://10.12.54.127:9200/]}}
[2018-05-17T15:17:40,179][INFO ][logstash.outputs.elasticsearch] Running health
check to see if an Elasticsearch connection is working {:healthcheck_url=>http:/
/10.12.54.127:9200/, :path=>"/"}
[2018-05-17T15:17:40,366][WARN ][logstash.outputs.elasticsearch] Restored connec
tion to ES instance {:url=>"http://10.12.54.127:9200/"}
[2018-05-17T15:17:40,425][INFO ][logstash.outputs.elasticsearch] ES Output versi
on determined {:es_version=>6}
[2018-05-17T15:17:40,430][WARN ][logstash.outputs.elasticsearch] Detected a 6.x
and above cluster: the `type` event field won't be used to determine the documen
t _type {:es_version=>6 }
[2018 -05 -17 T15:17 :40 ,445 ][INFO ][logstash.outputs.elasticsearch] Using mapping t
emplate from {:path=>nil}
[2018 -05 -17 T15:17 :40 ,462 ][INFO ][logstash.outputs.elasticsearch] Attempting to i
nstall template {:manage_template=>{"template" =>"logstash-*" , "version" =>60001 ,
"settings" =>{"index.refresh_interval" =>"5s" }, "mappings" =>{"_default_" =>{"dynami
c_templates"=>[{"message_field" =>{"path_match" =>"message" , "match_mapping_type" =
>"string" , "mapping" =>{"type" =>"text" , "norms" =>false}}}, {"string_fields" =>{"ma
tch"=>"*" , "match_mapping_type" =>"string" , "mapping" =>{"type" =>"text" , "norms" =>
false, "fields" =>{"keyword" =>{"type" =>"keyword" , "ignore_above" =>256 }}}}}], "pro
perties"=>{"@timestamp" =>{"type" =>"date" }, "@version" =>{"type" =>"keyword" }, "geo
ip"=>{"dynamic" =>true, "properties" =>{"ip" =>{"type" =>"ip" }, "location" =>{"type" =
>"geo_point" }, "latitude" =>{"type" =>"half_float" }, "longitude" =>{"type" =>"half_f
loat"}}}}}}}}
[2018 -05 -17 T15:17 :40 ,502 ][INFO ][logstash.outputs.elasticsearch] New Elasticsear
ch output {:class=>"LogStash::Outputs::ElasticSearch" , :hosts=>["http://10.12.54
.127:9200"]}
[2018 -05 -17 T15:17 :41 ,094 ][INFO ][logstash.pipeline ] Pipeline started suc
cesfully {:pipeline_id=>"main" , :thread=>"#<Thread:0x31bffa29 run>" }
[2018 -05 -17 T15:17 :41 ,199 ][INFO ][logstash.agent ] Pipelines running {:
count=>1 , :pipelines=>["main" ]}
2018 -05 -17 T07:19 :13.779 Z PC201801151216 SDFSDFSD
2018 -05 -17 T07:19 :13.781 Z PC201801151216 SDFSDF
2018 -05 -17 T07:19 :13.781 Z PC201801151216 SDFSD
2018 -05 -17 T07:19 :13.781 Z PC201801151216 SDFSDF
2018 -05 -17 T07:19 :13.781 Z PC201801151216 SDFSDF
2018 -05 -17 T07:19 :13.745 Z PC201801151216 TEST123
2018 -05 -17 T07:19 :13.781 Z PC201801151216 SDFSDF
Kibana中查看数据
在 Management
> Index Patterns
> Create Index Pattern
> Next step
选择 @timestamp
> Create index pattern
> Discover
可以看到我们测试的数据已经在Kibana
中了。
参考
【推荐】国内首个AI IDE,深度理解中文开发场景,立即下载体验Trae
【推荐】编程新体验,更懂你的AI,立即体验豆包MarsCode编程助手
【推荐】抖音旗下AI助手豆包,你的智能百科全书,全免费不限次数
【推荐】轻量又高性能的 SSH 工具 IShell:AI 加持,快人一步
· Linux系列:如何用 C#调用 C方法造成内存泄露
· AI与.NET技术实操系列(二):开始使用ML.NET
· 记一次.NET内存居高不下排查解决与启示
· 探究高空视频全景AR技术的实现原理
· 理解Rust引用及其生命周期标识(上)
· DeepSeek 开源周回顾「GitHub 热点速览」
· 物流快递公司核心技术能力-地址解析分单基础技术分享
· .NET 10首个预览版发布:重大改进与新特性概览!
· AI与.NET技术实操系列(二):开始使用ML.NET
· 单线程的Redis速度为什么快?