部署 elk 日志系统 elasticsearchlogstash kibana

Posted

tags:

篇首语:本文由小常识网(cha138.com)小编为大家整理,主要介绍了部署 elk 日志系统 elasticsearchlogstash kibana相关的知识,希望对你有一定的参考价值。

安装elk    
    安装Java 1.8环境
        解压源码安装包:

        tar xf jdk-8u121-linux-x64.tar.gz 
        ll
        mkdir /work/opt -p
        mv  jdk1.8.0_121  /work/opt/jdk
        ll /work/opt/jdk/
        chown -R root.root  /work/opt
         
        vim /etc/profile :    //添加
        export JAVA_HOME=/work/opt/jdk
        export JAVA_BIN=$JAVA_HOME/bin
        export JRE_HOME=${JAVA_HOME}/jre
        export CLASSPATH=$JAVA_HOME/jre/lib/ext:$JAVA_HOME/lib/tools.jar
        export PATH=$JAVA_BIN:$PATH

        
    安装 elasticsearch-5.3.0

    tar xf elasticsearch-5.3.0.tar.gz 
    mv elasticsearch-5.3.0 /work/opt/elasticsearch
    
    
    [email protected]:/work/opt/elasticsearch/config$ egrep -v ‘#|^$‘ elasticsearch.yml
    cluster.name: lvnian-elk
    node.name: lvnian-elk-node1
    path.data: /data  #由于是普通用户启动elasticsearch,所以这个目录的属主需要改为普通用户
    path.logs: /work/opt/elasticsearch/logs     #由于是普通用户启动elasticsearch,所以这个目录的属主需要改为普通用户
    bootstrap.memory_lock: false
    network.host: 0.0.0.0
    http.port: 9200
    http.cors.enabled: true 
    http.cors.allow-origin: "*" 
    [email protected]:/work/opt/elasticsearch/config$ 
    nohup /work/opt/elasticsearch/bin/elasticsearch >> /tmp/elasticsearch.log  &

   ##ES 5.1.1 安装 head:
    (5.1.1版本的 elasticsearch 没有提供直接插件安装方法,但在该github上该插件作者给出了方法)

        下载二进制源码包:
        wget https://nodejs.org/dist/v4.2.2/node-v6.2.0-linux-x64.tar.gz
        解压:
        tar xf node-v6.2.0-linux-x64.tar.gz -C /work/opt/
        设置环境变量:
        vim /etc/profile: 
            export NODE_HOME=/work/opt/node/
            export PATH=$PATH:$NODE_HOME/bin
         
        [email protected]:/work/source# node --version
        v6.10.1
    
        npm config set registry https://registry.npm.taobao.org    //设置代理镜像源,加速下载
        cd /home/stt/node-v4.2.2-linux-x64/lib/node_modules
        npm install grunt    //显示的2条warn可以忽略

        测试grunt是否生效:

        $ grunt -version
        grunt-cli v1.2.0
        grunt v1.0.1

        安装head插件:

          下载:git clone git://github.com/mobz/elasticsearch-head.git
        cd /home/stt/elasticsearch-head
        npm install (提示:如果遇到网络的瓶颈,将预先下载的源码包放在对应的位置效果一样,目录为/tmp/phantomjs/,确定位置后可自行创建并上传包)
        修改 elasticsearch-head/ _site/app.js
        // 把localhost改为ip
        找到:
        this.base_uri = this.config.base_uri || this.prefs.get("app-base_uri") || "http://localhost:9200";
        改为:
        this.base_uri = this.config.base_uri || this.prefs.get("app-base_uri") || "http://192.168.8.116:9200";
        修改 elasticsearch-head/Gruntfile.js
        connect: {
            server: {
                options: {
                    hostname: "0.0.0.0",     //增加这个配置
                    port: 9100,
                    base: ‘.‘,
                    keepalive: true
                }
            }
        }

        启动服务:(后台运行)
        grunt server &    //需要在 /home/stt/elasticsearch-head 下执行,因为我的 grunt 没有进行全局的安装


        
    ##安装Logstash(这个软件在你需要读取的日志服务上安装,用logstash读取你的日志,上传给elasticsearch的):
        也需要安装java1.8环境

        tar xf logstash-5.3.0.tar.gz 
        mkdir /work/opt
        mv logstash-5.3.0 /work/opt/
        cd /work/opt/

    vim /work/opt/logstash-5.3.0/conf/central.conf    #(处理基于 FILE 方式输入的日志信息,这里是简单的举个例子,日后继续学习补充)

        input {
            file {
                path => "/tmp/*.log"
            }
        }
        output {
            elasticsearch {
                hosts => "192.168.8.116:9200"
                index => "nginx-access"
            }
            stdout {
                codec => rubydebug
            }
        }

    
    
    
    ##安装Kibana:
    解压源码包:

    tar zxf kibana-5.1.1-linux-x86_64.tar.gz -C /home/stt/server/
    vim config/kibana.yml    //修改
        server.port: 5601    //打开注释而已,不用可以去效果,请使用默认端口
        server.host: "0.0.0.0"    //打开监听地址,让别的机器也能访问这个 kibana
        elasticsearch.url: "http://127.0.0.1:9200"      //这个url要根据实质情况,添加访问 elasticsearch 的url
    启动服务: (后台运行)
    /home/stt/server/kibana-5.1.1-linux-x86_64/bin/kibana &

        
    安装nginx 反向代理,
    apt-gei install nginx


    nginx放心代理配置文件如下:

 ##文件名kibana.conf
    upstream backend {
                 server 172.31.6.155:5601;
             }
      server {
            listen       80;
            server_name  kibana.lvnian.co;
        access_log /tmp/kibana-access.log;
        error_log /tmp/kibana-error.log;
            location / {
                #设置主机头和客户端真实地址,以便服务器获取客户端真实IP
                 proxy_set_header Host $host;
                 proxy_set_header X-Real-IP $remote_addr;
                 proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
                 #禁用缓存
                 proxy_buffering off;
                 #反向代理的地址
                 proxy_pass http://backend;     
            }
        }


        
        
        
    logstash 读取nginx的访问日志以及error日志,上传到logstash的配置文件文件。

   用下面命令运行即可

nohup /work/opt/logstash-5.3.0/bin/logstash -f /work/opt/logstash-5.3.0/conf/elk-nginx-log.conf &

   文件名: elk-nginx-log.conf
    input {
        file {
            path => "/data/log/nginx/*.log"
            start_position => beginning
        }
    }
    filter {
        if [path] =~ "access"  {
            mutate { replace => { type => "nginx_access" } }
            ruby {
                init => "@kname = [‘http_x_forwarded_for‘,‘time_local‘,‘request‘,‘status‘,‘body_bytes_sent‘,‘request_body‘,‘content_length‘,‘http_referer‘,‘http_user_agent‘,‘http_cookie‘,‘remote_addr‘,‘hostname‘,‘upstream_addr‘,‘upstream_response_time‘,‘request_time‘]"
                code => "
                    new_event = LogStash::Event.new(Hash[@kname.zip(event.get(‘message‘).split(‘|‘))])
                    new_event.remove(‘@timestamp‘)
                    event.append(new_event)
                "
            }
            if [request] {
                ruby {
                    init => "@kname = [‘method‘,‘uri‘,‘verb‘]"
                    code => "
                        new_event = LogStash::Event.new(Hash[@kname.zip(event.get(‘request‘).split(‘ ‘))])
                        new_event.remove(‘@timestamp‘)
                        event.append(new_event)
                    "
                }
                if [uri] {
                    ruby {
                        init => "@kname = [‘url_path‘,‘url_args‘]"
                        code => "
                            new_event = LogStash::Event.new(Hash[@kname.zip(event.get(‘uri‘).split(‘?‘))])
                            new_event.remove(‘@timestamp‘)
                            event.append(new_event)
                        "
                    }
                    kv {
                        prefix => "url_"
                        source => "url_args"
                        field_split => "& "
                        remove_field => [ "url_args","uri","request" ]
                    }
                }
            }
            mutate {
                convert => [
                    "body_bytes_sent" , "integer",
                    "content_length", "integer",
                    "upstream_response_time", "float",
                    "request_time", "float"
                ]
            }
            date {
                match => [ "time_local", "dd/MMM/yyyy:hh:mm:ss Z" ]
                locale => "en"
            }
        }
        else if [path] =~ "error" {
            mutate { replace => { type => "nginx_error" } }
            grok {
                match => { "message" => "(?<datetime>\d\d\d\d/\d\d/\d\d \d\d:\d\d:\d\d) \[(?<errtype>\w+)\] \S+: \*\d+ (?<errmsg>[^,]+), (?<errinfo>.*)$" }
            }
            mutate {
                rename => [ "host", "fromhost" ]
                gsub => [ "errmsg", "too large body: \d+ bytes", "too large body" ]
            }
            if [errinfo]
            {
                ruby {
                    code => "
                        new_event = LogStash::Event.new(Hash[event.get(‘errinfo‘).split(‘, ‘).map{|l| l.split(‘: ‘)}])
                        new_event.remove(‘@timestamp‘)
                        event.append(new_event)
                    "
                }
            }
            grok {
        #       match => { "request" => ‘"%{WORD:verb} %{URIPATH:urlpath}(?:\?%{NGX_URIPARAM:urlparam})?(?: HTTP/%{NUMBER:httpversion})"‘ }
                match => { "request" => ‘"%{WORD:verb} %{URIPATH:urlpath}?(?: HTTP/%{NUMBER:httpversion})"‘ }
                patterns_dir => ["/etc/logstash/patterns"]
        #      remove_field => [ "message", "errinfo", "request" ]
            }
        }
        else {
                mutate { replace => { type => "random_logs" } }
        }
    }
    output {
        elasticsearch {
            hosts => "172.31.1.79:9200"
            #index => "logstash-nginx"
            index => "logstash-%{type}-%{+YYYY.MM.dd}"
            document_type => "%{type}"
            flush_size => 20000
            idle_flush_time => 10
            sniffing => true
            template_overwrite => true
        }
        #    stdout {
        #        codec => rubydebug
        #    }
    }


本文出自 “奋斗吧” 博客,请务必保留此出处http://lvnian.blog.51cto.com/7155281/1915316

以上是关于部署 elk 日志系统 elasticsearchlogstash kibana的主要内容,如果未能解决你的问题,请参考以下文章

zabbix监控ELK导入的项目日志

Elk 进阶部署

ELK——ELK日志分析系统部署搭建

ELK企业级日志分析系统概述及部署

ELK日志分析系统(理论+部署)

ELK 企业级日志分析系统