码迷,mamicode.com
首页 > 其他好文 > 详细

集群介绍,keepalived介绍,用keepalived配置高可用集群

时间:2018-07-04 01:07:27      阅读:195      评论:0      收藏:0      [点我收藏+]

标签:业务   serve   ip add   http   测试   var   proxy   邮件发送   tail   

集群介绍
  • 根据功能划分为两大类:高可用和负载均衡
  • 高可用集群通常为两台服务器,一台工作,另外一台作为冗余,当提供服务的机器宕机,冗余将接替继续提供服务
  • 业内有个综合评分比如4个九,甚至5个九六个九来衡量服务的高可用,一般大企业的核心业务都有高可用,如果一个机器故障,一分钟之内就可以切换到另一台上
  • 实现高可用的开源软件有:heartbeat、keepalived
  • 负载均衡集群,需要有一台服务器作为分发器,它负责把用户的请求分发给后端的服务器处理,在这个集群里,除了分发器外,就是给用户提供服务的服务器了,这些服务器数量至少为2
  • 实现负载均衡的开源软件有LVS、keepalived、haproxy、nginx,商业的有F5、Netscaler

    keepalived介绍

  • 在这里我们使用keepalived来实现高可用集群,因为heartbeat在centos6上有一些问题,影响实验效果,并更新不及时
  • keepalived通过VRRP(Virtual Router Redundancy Protocl即虚拟路由冗余协议)来实现高可用。
  • 在这个协议里会将多台功能相同的路由器(其实是一台机器)组成一个小组,这个小组里会有1个master角色和N(N>=1)个backup角色。
  • master会通过组播的形式向各个backup发送VRRP协议的数据包,当backup收不到master发来的VRRP数据包时,就会认为master宕机了。此时就需要根据各个backup的优先级来决定谁成为新的mater。
  • Keepalived要有三个模块,分别是core、check和vrrp。其中core模块为keepalived的核心,负责主进程的启动、维护以及全局配置文件的加载和解析,check模块负责健康检查,vrrp模块是来实现VRRP协议的。

    用keepalived配置高可用集群

  • 准备两台机器130和132,130作为master,132作为backup
  • 两台机器都执行yum install -y keepalived
  • 两台机器都安装nginx,其中130上已经编译安装过nginx,132上需要yum安装nginx: yum install -y nginx
    • 之所以选择nginx,在生产环境中,好多企业把它作为负载均衡器,如果它挂掉,影响后面好多web服务器,所以不能出现单点故障
  • 编辑130上的keepalived配置文件,
    [root@akuilinux01 ~]# > /etc/keepalived/keepalived.conf 
    [root@akuilinux01 ~]# vim !$
    vim /etc/keepalived/keepalived.conf
    global_defs {                      #这个全局定义参数
    notification_email {            #出现问题给邮箱发邮件,
     aming@aminglinux.com
    }
    notification_email_from root@aminglinux.com #由哪一个邮件发送
    smtp_server 127.0.0.1
    smtp_connect_timeout 30
    router_id LVS_DEVEL
    }
    vrrp_script chk_nginx {        #检测一个服务是否正常,需要有个脚本
    script "/usr/local/sbin/check_ng.sh"
    interval 3   # 3秒检测一次
    }
    vrrp_instance VI_1 {   # 定义master相关的
    state MASTER       #如果是从就是backup    
    interface ens33    #定义网卡 
    virtual_router_id 51  # 定义路由器的id
    priority 100          #定义权重
    advert_int 1          
    authentication {        #认证相关    
        auth_type PASS
        auth_pass aminglinux>com
    }
    virtual_ipaddress {   #定义vip(公有的ip)
        192.168.21.100
    }
    track_script {    #加载服务
        chk_nginx
    }
    }
  • 在130上编辑监控脚本
    [root@akuilinux01 ~]# vim /usr/local/sbin/check_ng.sh
    #!/bin/bash
    #时间变量,用于记录日志
    d=`date --date today +%Y%m%d_%H:%M:%S`
    #计算nginx进程数量
    n=`ps -C nginx --no-heading|wc -l`
    #如果进程为0,则启动nginx,并且再次检测nginx进程数量,
    #如果还为0,说明nginx无法启动,此时需要关闭keepalived(防止脑裂)
    if [ $n -eq "0" ]; then
        /etc/init.d/nginx start
        n2=`ps -C nginx --no-heading|wc -l`
        if [ $n2 -eq "0"  ]; then
                echo "$d nginx down,keepalived will stop" >> /var/log/check_ng.log
                systemctl stop keepalived
        fi
    fi
  • 给脚本755权限,并启动keepalived服务
    [root@akuilinux01 ~]# chmod 755 /usr/local/sbin/check_ng.sh 
    [root@akuilinux01 ~]# systemctl start keepalived
    [root@akuilinux01 ~]# ps aux |grep keepa
    root      2895  0.0  0.0 118608  1388 ?        Ss   23:20   0:00 /usr/sbin/keepalived -D
    root      2896  0.1  0.1 127468  3304 ?        S    23:20   0:00 /usr/sbin/keepalived -D
    root      2899  0.3  0.1 127408  2836 ?        S    23:20   0:00 /usr/sbin/keepalived -D
    root      2961  0.0  0.0 112676   980 pts/1    S+   23:20   0:00 grep --color=auto keepa
  • 检测下nginx可以自动加载不
    [root@akuilinux01 ~]# ps aux |grep nginx
    root      1016  0.0  0.0  46008  1260 ?        Ss   20:52   0:00 nginx: master process /usr/localnginx/sbin/nginx -c /usr/local/nginx/conf/nginx.conf
    nobody    1017  0.0  0.2  48496  4152 ?        S    20:52   0:00 nginx: worker process
    nobody    1018  0.0  0.2  48496  3896 ?        S    20:52   0:00 nginx: worker process
    root      2982  0.0  0.0 112680   980 pts/1    S+   23:20   0:00 grep --color=auto nginx
    [root@akuilinux01 ~]# /etc/init.d/nginx stop
    Stopping nginx (via systemctl):                            [  确定  ]
    [root@akuilinux01 ~]# ps aux |grep nginx
    root      3266  1.0  0.0  46008  1252 ?        Ss   23:22   0:00 nginx: master process /usr/localnginx/sbin/nginx -c /usr/local/nginx/conf/nginx.conf
    nobody    3267  0.0  0.2  48496  3892 ?        S    23:22   0:00 nginx: worker process
    nobody    3268  0.0  0.2  48496  3892 ?        S    23:22   0:00 nginx: worker process
    root      3273  0.0  0.0 112676   984 pts/1    S+   23:22   0:00 grep --color=auto nginx
  • 日志
    [root@akuilinux01 ~]# less /var/log/messages
  • 查看vip
    [root@akuilinux01 ~]# ip add
    1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN qlen 1
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
    2: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:0c:29:87:fb:87 brd ff:ff:ff:ff:ff:ff
    inet 192.168.21.128/24 brd 192.168.21.255 scope global ens33
       valid_lft forever preferred_lft forever
    inet 192.168.21.100/32 scope global ens33
       valid_lft forever preferred_lft forever
    inet6 fe80::68fd:e6fa:c781:f5a6/64 scope link 
       valid_lft forever preferred_lft forever
    3: ens37: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:0c:29:87:fb:91 brd ff:ff:ff:ff:ff:ff
    inet 192.168.110.136/24 brd 192.168.110.255 scope global dynamic ens37
       valid_lft 1278sec preferred_lft 1278sec
    inet6 fe80::c559:4a92:72f1:b448/64 scope link 
       valid_lft forever preferred_lft forever
  • 132上编辑配置文件

    [root@akuilinux02 ~]# > /etc/keepalived/keepalived.conf 
    [root@akuilinux02 ~]# vim /etc/keepalived/keepalived.conf 
    global_defs {
    notification_email {
     aming@aminglinux.com
    }
    notification_email_from root@aminglinux.com
    smtp_server 127.0.0.1
    smtp_connect_timeout 30
    router_id LVS_DEVEL
    }
    vrrp_script chk_nginx {
    script "/usr/local/sbin/check_ng.sh"
    interval 3
    }
    vrrp_instance VI_1 {
    state BACKUP
    interface ens33
    virtual_router_id 51
    priority 90
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass aminglinux>com
    }
    virtual_ipaddress {
        192.168.21.100
    }
    
    track_script {
        chk_nginx
    }
    }
  • 132上编辑监控脚本
    [root@akuilinux02 ~]# vim /usr/local/sbin/check_ng.sh
    #时间变量,用于记录日志
    d=`date --date today +%Y%m%d_%H:%M:%S`
    #计算nginx进程数量
    n=`ps -C nginx --no-heading|wc -l`
    #如果进程为0,则启动nginx,并且再次检测nginx进程数量,
    #如果还为0,说明nginx无法启动,此时需要关闭keepalived
    if [ $n -eq "0" ]; then
        systemctl start nginx
        n2=`ps -C nginx --no-heading|wc -l`
        if [ $n2 -eq "0"  ]; then
                echo "$d nginx down,keepalived will stop" >> /var/log/check_ng.log
                systemctl stop keepalived
        fi
    fi
  • 给脚本755权限并启动keepalived
    [root@akuilinux02 ~]# chmod 755 /usr/local/sbin/check_ng.sh
    [root@akuilinux02 ~]# systemctl start keepalived

    测试

  • curl或者浏览器打开
    [root@akuilinux01 ~]# curl localhost@192.168.21.128 
    master,master.
    [root@akuilinux01 ~]# curl  localhost@192.168.21.129
    backup,backup.
    [root@akuilinux01 ~]# curl  localhost@192.168.21.100
    master,master.
  • 测试高可用
    [root@akuilinux01 ~]# systemctl stop keepalived
    [root@akuilinux01 ~]# curl localhost@192.168.21.100
    backup,backup.
    [root@akuilinux01 ~]# systemctl start  keepalived
    [root@akuilinux01 ~]# curl localhost@192.168.21.100
    master,master.

    扩展

  • heartbeat和keepalived比较
  • DRBD工作原理和配置
  • mysql+keepalived

集群介绍,keepalived介绍,用keepalived配置高可用集群

标签:业务   serve   ip add   http   测试   var   proxy   邮件发送   tail   

原文地址:http://blog.51cto.com/akui2521/2135810

(0)
(0)
   
举报
评论 一句话评论(0
登录后才能评论!
© 2014 mamicode.com 版权所有  联系我们:gaon5@hotmail.com
迷上了代码!