Centos7安装ELK开源日志应用

一、介绍

　　

　　The Elastic Stack - 它不是一个软件，而是Elasticsearch，Logstash，Kibana 开源软件的集合，对外是作为一个日志管理系统的开源方案。它可以从任何来源，任何格式进行日志搜索，分析获取数据，并实时进行展示。像盾牌（安全），监护者（警报）和Marvel（监测）一样为你的产品提供更多的可能。

　　Elasticsearch：搜索，提供分布式全文搜索引擎

　　Logstash: 日志收集，管理，存储

　　Kibana ：日志的过滤web 展示
　　Filebeat：监控日志文件、转发

 二、测试环境规划图

　　　　

　　环境：ip、主机名按照如上规划，系统已经 update. 所有主机时间一致。防火墙测试环境已关闭。下面是这次elk学习的部署安装

　　目的：通过elk 主机收集监控主要server的系统日志、以及线上应用服务日志。

三、Elasticsearch+Logstash+Kibana的安装（在 elk.test.com 上进行操作）

3.1.基础环境检查

[root@elk ~]# hostnameelk.test.com
[root@elk ~]# cat /etc/hosts127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4
::1         localhost localhost.localdomain localhost6 localhost6.localdomain6192.168.30.67   elk.test.com192.168.30.99   rsyslog.test.com192.168.30.64   nginx.test.com

3.2.软件包

[root@elk ~]# cd elk/[root@elk elk]# wget -c https://download.elastic.co/elasticsearch/release/org/elasticsearch/distribution/rpm/elasticsearch/2.3.3/elasticsearch-2.3.3.rpm[root@elk elk]# wget -c https://download.elastic.co/logstash/logstash/packages/centos/logstash-2.3.2-1.noarch.rpm[root@elk elk]# wget https://download.elastic.co/kibana/kibana/kibana-4.5.1-1.x86_64.rpm[root@elk elk]# wget -c https://download.elastic.co/beats/filebeat/filebeat-1.2.3-x86_64.rpm

3.3.检查

[root@elk elk]# lselasticsearch-2.3.3.rpm  filebeat-1.2.3-x86_64.rpm  kibana-4.5.1-1.x86_64.rpm  logstash-2.3.2-1.noarch.rpm

服务器只需要安装e、l、k, 客户端只需要安装filebeat。

 3.4.安装elasticsearch，先安装jdk，elk server 需要java 开发环境支持，由于客户端上使用的是filebeat软件，它不依赖java环境，所以不需要安装。

[root@elk elk]# yum install java-1.8.0-openjdk -y

安装es

 localinstall elasticsearch-..rpm --.-.noarch                                                                                             / systemctl daemon--.-.noarch                                                                                             /:.-

重新载入 systemd，扫描新的或有变动的单元；启动并加入开机自启动

[root@elk elk]# systemctl daemon-reload
[root@elk elk]# systemctl enable elasticsearch
Created symlink from /etc/systemd/system/multi-user.target.wants/elasticsearch.service to /usr/lib/systemd/system/elasticsearch.service.
[root@elk elk]# systemctl start elasticsearch
[root@elk elk]# systemctl status elasticsearch
● elasticsearch.service - Elasticsearch
   Loaded: loaded (/usr/lib/systemd/system/elasticsearch.service; enabled; vendor preset: disabled)
   Active: active (running) since Fri 2016-05-20 15:38:35 CST; 12s ago
     Docs: http://www.elastic.co
  Process: 10428 ExecStartPre=/usr/share/elasticsearch/bin/elasticsearch-systemd-pre-exec (code=exited, status=0/SUCCESS)
 Main PID: 10430 (java)
   CGroup: /system.slice/elasticsearch.service
           └─10430 /bin/java -Xms256m -Xmx1g -Djava.awt.headless=true -XX:+UseParNewGC -XX:+UseConcMarkSweepGC -XX:CMSInitiatingOccupancy...

May 20 15:38:38 elk.test.com elasticsearch[10430]: [2016-05-20 15:38:38,279][INFO ][env                      ] [James Howlett] heap...[true]
May 20 15:38:38 elk.test.com elasticsearch[10430]: [2016-05-20 15:38:38,279][WARN ][env                      ] [James Howlett] max ...65536]
May 20 15:38:41 elk.test.com elasticsearch[10430]: [2016-05-20 15:38:41,726][INFO ][node                     ] [James Howlett] initialized
May 20 15:38:41 elk.test.com elasticsearch[10430]: [2016-05-20 15:38:41,726][INFO ][node                     ] [James Howlett] starting ...
May 20 15:38:41 elk.test.com elasticsearch[10430]: [2016-05-20 15:38:41,915][INFO ][transport                ] [James Howlett] publ...:9300}
May 20 15:38:41 elk.test.com elasticsearch[10430]: [2016-05-20 15:38:41,920][INFO ][discovery                ] [James Howlett] elas...xx35hw
May 20 15:38:45 elk.test.com elasticsearch[10430]: [2016-05-20 15:38:45,099][INFO ][cluster.service          ] [James Howlett] new_...eived)
May 20 15:38:45 elk.test.com elasticsearch[10430]: [2016-05-20 15:38:45,164][INFO ][gateway                  ] [James Howlett] reco..._state
May 20 15:38:45 elk.test.com elasticsearch[10430]: [2016-05-20 15:38:45,185][INFO ][http                     ] [James Howlett] publ...:9200}
May 20 15:38:45 elk.test.com elasticsearch[10430]: [2016-05-20 15:38:45,185][INFO ][node                     ] [James Howlett] started
Hint: Some lines were ellipsized, use -l to show in full.

 

 检查服务

[root@elk elk]# rpm -qc elasticsearch/etc/elasticsearch/elasticsearch.yml/etc/elasticsearch/logging.yml/etc/init.d/elasticsearch/etc/sysconfig/elasticsearch/usr/lib/sysctl.d/elasticsearch.conf/usr/lib/systemd/system/elasticsearch.service/usr/lib/tmpfiles.d/elasticsearch.conf
[root@elk elk]# netstat -nltp | grep java
tcp6       0      0 127.0.0.1:9200          :::*                    LISTEN      10430/java
tcp6       0      0 ::1:9200                :::*                    LISTEN      10430/java
tcp6       0      0 127.0.0.1:9300          :::*                    LISTEN      10430/java
tcp6       0      0 ::1:9300                :::*                    LISTEN      10430/java

修改防火墙，将9200、9300 端口对外开放

[root@elk elk]# firewall-cmd --permanent --add-port={9200/tcp,9300/tcp}
success
[root@elk elk]# firewall-cmd --reload
success
[root@elk elk]# firewall-cmd  --list-all
public (default, active)
  interfaces: eno16777984 eno33557248
  sources:
  services: dhcpv6-client ssh
  ports: 9200/tcp 9300/tcp
  masquerade: no
  forward-ports:
  icmp-blocks:
  rich rules:

3.5 安装kibana

[root@elk elk]# yum localinstall kibana-4.5.1-1.x86_64.rpm –y
[root@elk elk]# systemctl enable kibana
Created symlink from /etc/systemd/system/multi-user.target.wants/kibana.service to /usr/lib/systemd/system/kibana.service.
[root@elk elk]# systemctl start kibana

[root@elk elk]# systemctl status kibana
● kibana.service - no description given
   Loaded: loaded (/usr/lib/systemd/system/kibana.service; enabled; vendor preset: disabled)
   Active: active (running) since Fri 2016-05-20 15:49:02 CST; 20s ago
 Main PID: 11260 (node)
   CGroup: /system.slice/kibana.service
           └─11260 /opt/kibana/bin/../node/bin/node /opt/kibana/bin/../src/cli

May 20 15:49:05 elk.test.com kibana[11260]: {"type":"log","@timestamp":"2016-05-20T07:49:05+00:00","tags":["status","plugin:elasticsearch...May 20 15:49:05 elk.test.com kibana[11260]: {"type":"log","@timestamp":"2016-05-20T07:49:05+00:00","tags":["status","plugin:kbn_vi...lized"}
May 20 15:49:05 elk.test.com kibana[11260]: {"type":"log","@timestamp":"2016-05-20T07:49:05+00:00","tags":["status","plugin:markdo...lized"}
May 20 15:49:05 elk.test.com kibana[11260]: {"type":"log","@timestamp":"2016-05-20T07:49:05+00:00","tags":["status","plugin:metric...lized"}
May 20 15:49:05 elk.test.com kibana[11260]: {"type":"log","@timestamp":"2016-05-20T07:49:05+00:00","tags":["status","plugin:spyMod...lized"}
May 20 15:49:05 elk.test.com kibana[11260]: {"type":"log","@timestamp":"2016-05-20T07:49:05+00:00","tags":["status","plugin:status...lized"}
May 20 15:49:05 elk.test.com kibana[11260]: {"type":"log","@timestamp":"2016-05-20T07:49:05+00:00","tags":["status","plugin:table_...lized"}
May 20 15:49:05 elk.test.com kibana[11260]: {"type":"log","@timestamp":"2016-05-20T07:49:05+00:00","tags":["listening","info"],"pi...:5601"}
May 20 15:49:10 elk.test.com kibana[11260]: {"type":"log","@timestamp":"2016-05-20T07:49:10+00:00","tags":["status","plugin:elasticsearch...May 20 15:49:14 elk.test.com kibana[11260]: {"type":"log","@timestamp":"2016-05-20T07:49:14+00:00","tags":["status","plugin:elasti...found"}
Hint: Some lines were ellipsized, use -l to show in full.

检查kibana服务运行（Kibana默认 进程名：node ，端口5601）

[root@elk elk]# netstat -nltp
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address           Foreign Address         State       PID/Program name
tcp        0      0 0.0.0.0:22              0.0.0.0:*               LISTEN      909/sshd
tcp        0      0 127.0.0.1:25            0.0.0.0:*               LISTEN      1595/master
tcp        0      0 0.0.0.0:5601            0.0.0.0:*               LISTEN      11260/node

修改防火墙，对外开放tcp/5601

[root@elk elk]# firewall-cmd --permanent --add-port=5601/tcp
Success
[root@elk elk]# firewall-cmd --reload
success
[root@elk elk]# firewall-cmd --list-all
public (default, active)
  interfaces: eno16777984 eno33557248
  sources:
  services: dhcpv6-client ssh
  ports: 9200/tcp 9300/tcp 5601/tcp
  masquerade: no
  forward-ports:
  icmp-blocks:
  rich rules:

这时，我们可以打开浏览器，测试访问一下kibana服务器http://192.168.30.67:5601/，确认没有问题，如下图：

在这里，我们可以修改防火墙，将用户访问80端口连接转发到5601上，这样可以直接输入网址不用指定端口了，如下:

[root@elk elk]# firewall-cmd --permanent --add-forward-port=port=80:proto=tcp:toport=5601[root@elk elk]# firewall-cmd --reload[root@elk elk]# firewall-cmd --list-all
public (default, active)
  interfaces: eno16777984 eno33557248
  sources:
  services: dhcpv6-client ssh
  ports: 9200/tcp 9300/tcp 5601/tcp
  masquerade: no
  forward-ports: port=80:proto=tcp:toport=5601:toaddr=
  icmp-blocks:
  rich rules:

3.6 安装logstash，以及添加配置文件

[root@elk elk]# yum localinstall logstash-2.3.2-1.noarch.rpm –y

生成证书

[root@elk elk]# cd /etc/pki/tls/[root@elk tls]# lscert.pem  certs  misc  openssl.cnf  private

[root@elk tls]# openssl req -subj '/CN=elk.test.com/' -x509 -days 3650 -batch -nodes -newkey rsa:2048 -keyout private/logstash-forwarder.key -out 
certs/logstash-forwarder.crt
Generating a 2048 bit RSA private key
...................................................................+++......................................................+++writing new private key to 'private/logstash-forwarder.key'-----

之后创建logstash 的配置文件。如下：

[root@elk ~]# cat /etc/logstash/conf.d/01-logstash-initial.conf
input {
  beats {
    port => 5000
    type => "logs"
    ssl => true
    ssl_certificate => "/etc/pki/tls/certs/logstash-forwarder.crt"
    ssl_key => "/etc/pki/tls/private/logstash-forwarder.key"
  }
}

filter {  if [type] == "syslog-beat" {
    grok {
      match => { "message" => "%{SYSLOGTIMESTAMP:syslog_timestamp} %{SYSLOGHOST:syslog_hostname} %{DATA:syslog_program}(?:\[%{POSINT:syslog_pid}\])?: %{GREEDYDATA:syslog_message}" }
      add_field => [ "received_at", "%{@timestamp}" ]
      add_field => [ "received_from", "%{host}" ]
    }
    geoip {
      source => "clientip"
    }
    syslog_pri {}    date {
      match => [ "syslog_timestamp", "MMM d HH:mm:ss", "MMM dd HH:mm:ss" ]
    }
  }
}

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

启动logstash，并检查端口，配置文件里，我们写的是5000端口

[root@elk conf.d]# systemctl start logstash
[root@elk elk]# /sbin/chkconfig logstash on
[root@elk conf.d]# netstat -ntlp
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address           Foreign Address         State       PID/Program name
tcp        0      0 0.0.0.0:22              0.0.0.0:*               LISTEN      909/sshd
tcp        0      0 127.0.0.1:25            0.0.0.0:*               LISTEN      1595/master
tcp        0      0 0.0.0.0:5601            0.0.0.0:*               LISTEN      11260/node
tcp        0      0 0.0.0.0:514             0.0.0.0:*               LISTEN      618/rsyslogd
tcp6       0      0 :::5000                 :::*                    LISTEN      12819/java
tcp6       0      0 :::3306                 :::*                    LISTEN      1270/mysqld
tcp6       0      0 127.0.0.1:9200          :::*                    LISTEN      10430/java
tcp6       0      0 ::1:9200                :::*                    LISTEN      10430/java
tcp6       0      0 127.0.0.1:9300          :::*                    LISTEN      10430/java
tcp6       0      0 ::1:9300                :::*                    LISTEN      10430/java
tcp6       0      0 :::22                   :::*                    LISTEN      909/sshd
tcp6       0      0 ::1:25                  :::*                    LISTEN      1595/master
tcp6       0      0 :::514                  :::*                    LISTEN      618/rsyslogd

 

修改防火墙，将5000端口对外开放。

[root@elk ~]# firewall-cmd --permanent --add-port=5000/tcp
success
[root@elk ~]# firewall-cmd --reload
success
[root@elk ~]# firewall-cmd --list-all
public (default, active)
  interfaces: eno16777984 eno33557248
  sources:
  services: dhcpv6-client ssh
  ports: 9200/tcp 9300/tcp 5000/tcp 5601/tcp
  masquerade: no
  forward-ports: port=80:proto=tcp:toport=5601:toaddr=
  icmp-blocks:
  rich rules:

3.7 修改elasticsearch 配置文件

查看目录，创建文件夹es-01（名字不是必须的）,logging.yml是自带的，elasticsearch.yml是创建的文件，内如见下：

[root@elk ~]# cd /etc/elasticsearch/[root@elk elasticsearch]# tree
.
├── es-01│   ├── elasticsearch.yml
│   └── logging.yml
└── scripts

[root@elk elasticsearch]# cat es-01/elasticsearch.yml 
----http:
  port: 9200network:
  host: elk.test.com
node:
  name: elk.test.com
path:
  data: /etc/elasticsearch/data/es-01

3.8 重启elasticsearch、logstash服务。

3.9 将 fiebeat安装包拷贝到 rsyslog、nginx 客户端上

[root@elk elk]# scp filebeat-1.2.3-x86_64.rpm [email protected]:/root/elk
[root@elk elk]# scp filebeat-1.2.3-x86_64.rpm [email protected]:/root/elk
[root@elk elk]# scp /etc/pki/tls/certs/logstash-forwarder.crt rsyslog.test.com:/root/elk
[root@elk elk]# scp /etc/pki/tls/certs/logstash-forwarder.crt nginx.test.com:/root/elk

 

四、客户端部署filebeat（在rsyslog、nginx客户端上操作）

　　

　　filebeat客户端是一个轻量级的，从服务器上的文件收集日志资源的工具，这些日志转发到处理到Logstash服务器上。该Filebeat客户端使用安全的Beats协议与Logstash实例通信。lumberjack协议被设计为可靠性和低延迟。Filebeat使用托管源数据的计算机的计算资源，并且Beats输入插件尽量减少对Logstash的资源需求。

4.1.（node1）安装filebeat，拷贝证书，创建收集日志配置文件

[root@rsyslog elk]# yum localinstall filebeat-1.2.3-x86_64.rpm -y
#拷贝证书到本机指定目录中
[root@rsyslog elk]# cp logstash-forwarder.crt /etc/pki/tls/certs/.
[root@rsyslog elk]# cd /etc/filebeat/[root@rsyslog filebeat]# tree
.
├── conf.d
│   ├── authlogs.yml
│   └── syslogs.yml
├── filebeat.template.json
└── filebeat.yml1 directory, 4 files

修改的文件有3个，filebeat.yml，是定义连接logstash 服务器的配置。conf.d目录下的2个配置文件是自定义监控日志的，下面看下各自的内容：

filebeat.yml

[root@rsyslog filebeat]# cat filebeat.yml 
filebeat:
  spool_size: 1024
  idle_timeout: 5s
  registry_file: .filebeat
  config_dir: /etc/filebeat/conf.d
output:
  logstash:
    hosts:    - elk.test.com:5000
    tls:
      certificate_authorities: ["/etc/pki/tls/certs/logstash-forwarder.crt"]
    enabled: trueshipper: {}
logging: {}
runoptions: {}

authlogs.yml  & syslogs.yml

[root@rsyslog filebeat]# cat conf.d/authlogs.yml 
filebeat:
  prospectors:    - paths:      - /var/log/secure
      encoding: plain
      fields_under_root: false
      input_type: log
      ignore_older: 24h
      document_type: syslog-beat
      scan_frequency: 10s
      harvester_buffer_size: 16384
      tail_files: false
      force_close_files: false
      backoff: 1s
      max_backoff: 1s
      backoff_factor: 2
      partial_line_waiting: 5s
      max_bytes: 10485760[root@rsyslog filebeat]# cat conf.d/syslogs.yml                     
filebeat:
  prospectors:    - paths:      - /var/log/messages
      encoding: plain
      fields_under_root: false
      input_type: log
      ignore_older: 24h
      document_type: syslog-beat
      scan_frequency: 10s
      harvester_buffer_size: 16384
      tail_files: false
      force_close_files: false
      backoff: 1s
      max_backoff: 1s
      backoff_factor: 2
      partial_line_waiting: 5s
      max_bytes: 10485760

修改完成后，启动filebeat服务

[root@rsyslog filebeat]# service filebeat start
Starting filebeat:                                         [  OK  ]
[root@rsyslog filebeat]# chkconfig filebeat on

[root@rsyslog filebeat]# netstat -altp
Active Internet connections (servers and established)
Proto Recv-Q Send-Q Local Address               Foreign Address             State       PID/Program name   
tcp        0      0 localhost:25151             *:*                         LISTEN      6230/python2        
tcp        0      0 *:ssh                       *:*                         LISTEN      5509/sshd           
tcp        0      0 localhost:ipp               *:*                         LISTEN      1053/cupsd          
tcp        0      0 localhost:smtp              *:*                         LISTEN      1188/master         
tcp        0      0 rsyslog.test.com:51155      elk.test.com:commplex-main  ESTABLISHED 7443/filebeat       
tcp        0     52 rsyslog.test.com:ssh        192.168.30.65:10580         ESTABLISHED 7164/sshd           
tcp        0      0 *:ssh                       *:*                         LISTEN      5509/sshd           
tcp        0      0 localhost:ipp               *:*                         LISTEN      1053/cupsd          
tcp        0      0 localhost:smtp              *:*                         LISTEN      1188/master

如果连接不上，状态不正常的话，检查下客户端的防火墙。

4.2. （node2）安装filebeat，拷贝证书，创建收集日志配置文件

[root@nginx elk]# yum localinstall filebeat-1.2.3-x86_64.rpm -y
[root@nginx elk]# cp logstash-forwarder.crt /etc/pki/tls/certs/.
[root@nginx elk]# cd /etc/filebeat/[root@nginx filebeat]# tree
.
├── conf.d
│   ├── nginx.yml
│   └── syslogs.yml
├── filebeat.template.json
└── filebeat.yml1 directory, 4 files

修改filebeat.yml 内容如下：

[root@rsyslog filebeat]# cat filebeat.yml 
filebeat:
  spool_size: 1024
  idle_timeout: 5s
  registry_file: .filebeat
  config_dir: /etc/filebeat/conf.d
output:
  logstash:
    hosts:    - elk.test.com:5000
    tls:
      certificate_authorities: ["/etc/pki/tls/certs/logstash-forwarder.crt"]
    enabled: trueshipper: {}
logging: {}
runoptions: {}

syslogs.yml & nginx.yml

[root@nginx filebeat]# cat conf.d/syslogs.yml 
filebeat:
  prospectors:    - paths:      - /var/log/messages
      encoding: plain
      fields_under_root: false
      input_type: log
      ignore_older: 24h
      document_type: syslog-beat
      scan_frequency: 10s
      harvester_buffer_size: 16384
      tail_files: false
      force_close_files: false
      backoff: 1s
      max_backoff: 1s
      backoff_factor: 2
      partial_line_waiting: 5s
      max_bytes: 10485760[root@nginx filebeat]# cat conf.d/nginx.yml 
filebeat:
  prospectors:    - paths:      - /var/log/nginx/access.log
      encoding: plain
      fields_under_root: false
      input_type: log
      ignore_older: 24h
      document_type: syslog-beat
      scan_frequency: 10s
      harvester_buffer_size: 16384
      tail_files: false
      force_close_files: false
      backoff: 1s
      max_backoff: 1s
      backoff_factor: 2
      partial_line_waiting: 5s
      max_bytes: 10485760

修改完成后，启动filebeat服务，并检查filebeat进程

[root@nginx filebeat]# service filebeat start
Starting filebeat:                                         [  OK  ]
[root@nginx filebeat]# chkconfig filebeat on

[root@nginx filebeat]# netstat -aulpt
Active Internet connections (servers and established)
Proto Recv-Q Send-Q Local Address               Foreign Address             State       PID/Program name   
tcp        0      0 *:ssh                       *:*                         LISTEN      1076/sshd           
tcp        0      0 localhost:smtp              *:*                         LISTEN      1155/master         
tcp        0      0 *:http                      *:*                         LISTEN      1446/nginx          
tcp        0     52 nginx.test.com:ssh          192.168.30.65:11690         ESTABLISHED 1313/sshd           
tcp        0      0 nginx.test.com:49500        elk.test.com:commplex-main  ESTABLISHED 1515/filebeat       
tcp        0      0 nginx.test.com:ssh          192.168.30.65:6215          ESTABLISHED 1196/sshd           
tcp        0      0 nginx.test.com:ssh          192.168.30.65:6216          ESTABLISHED 1200/sshd           
tcp        0      0 *:ssh                       *:*                         LISTEN      1076/sshd

 

通过上面可以看出，客户端filebeat进程已经和 elk 服务器连接了。下面去验证。

五、验证，访问kibana http://192.168.30.67

5.1 设置下

查看下两台机器的系统日志：node1的

node2的nginx 访问日志

 

六、体验

　　之前在学习rsyslog +LogAnalyzer，然后又学了这个之后，发现elk 不管从整体系统，还是体验都是不错的，而且更新快。后续会继续学习，更新相关的监控过滤日志方法，日志分析，以及使用kafka 来进行存储的架构。

 

本文章属于原创，如果觉得有价值，转载时请注明出处。谢谢

参考网站：https://www.elastic.co/products/elasticsearch

　　　　　https://www.elastic.co/downloads