keepalived之ipvs-dr

keepalived初试

Keepalived实现LVS-DR集群高可用

一、集群高可用概述

单纯的lvs/nginx反向代理模型做负载集群应用时,DR(director)存在单点故障隐患,故需要有机制来保证DR的高可用性。常用的高可用性方案有Keepalived、corosync,Keepalived主要是由VRRP协议实现了VIPfloating,比较适用于前端DR的高可用性,Corosync一般用于更专业的集群模型实现Service的高可用。Keepalived起初就是为了实现LVS集群director高可用而开发的,本处仅做Keepalived+LVS-DR模型实验。

二、Keepalived原理简介

Keepalived中优先级高的节点为MASTER。MASTER其中一个职责就是响应VIP的arp包,将VIP和mac地址映射关系告诉局域网内其 他主机,同时,它还会以多播的形式向局域网中发送VRRP通告,告知BACKUP组自己的优先级。网络中的所有BACKUP节点只负责 处理MASTER发出的多播包,当发现MASTER的优先级没自己高(脚本检测故障触发自我降级),或者没收到MASTER的VRRP通告(网络故障/MASTER宕机)时,BACKUP将自己切换到MASTER状 态,然后做MASTER该做的事:1.响应arp包,2.发送VRRP通告。

三、实验环境

1.网络拓补图

keepalived与lvs-dr

 

 2.软件环境

  • CentOS7.4

  • keepalived.x86_64                         1.3.5-6.el7

  • nginx.x86_64                              1:1.12.2-2.el7

四、配置流程

(1)两台DR配置keepalived.conf

  1. 配置DR1:

 global_defs {
                    notification_email {
                        root@localhost #此处仅发给本机,更定制化的邮件通知功能一般由zabbix来做。
                    }
                    notification_email_from keepalived@localhost
                    smtp_server 127.0.0.1
                    smtp_connect_timeout 30
                    router_id keepalivedR1  #定义路由器标识,每台服务器局域网内唯一就行。
                    vrrp_mcast_group4 224.0.0.33 #定义master向backup组播vrrp报文的地址。
                }
               vrrp_script chk_down {
                script "/etc/keepalived/chk_down.sh" 
                interval 1   #脚本检测间隔
                weight -15   #即原有优先级+weight,负值即降低。
# 注意当weight=0时, 用于变更vrrp_instance的状态例如脚本检测失败, 则vrrp的状态直接变为FAULT. (不管有没有其他节点存在来接管MASTER)
                fall 2       #执行脚本两次exit非0则降低优先级
                rise 1      #执行脚本一次exit 0 则还原优先级    
                user keepalived_script #默认用户yum安装keepalived需自建,不存在则调用root(不推荐)
               }
               
                vrrp_instance VI_1 {
                    state MASTER    #自定义的state,但如果你的优先级小于backup也不会成为master。
                    interface ens39       #绑定为当前虚拟路由器使用的物理接口。
                    virtual_router_id 3  #当前虚拟路由器的惟一标识,范围是0-255。同实例一致!
                    priority 99         #初始优先级,范围1-254。
                    advert_int 1        #vrrp通告的时间间隔。
                    authentication {
                        auth_type PASS   #简单密码验证,不超过8位。
                        auth_pass 736w4ib2  #最好使用随机字符串,同vip实例保持一致!
                    }
                    virtual_ipaddress {
                        192.168.7.120/24 dev ens39
                    }
                    notify_master "/etc/keepalived/notify.sh master" #调用通知脚本
                    notify_backup "/etc/keepalived/notify.sh backup"
                    notify_fault "/etc/keepalived/notify.sh fault"
                }
               virtual_server 192.168.7.120 80 {  #此处可用IP port/fwmark id 标识VIP对应服务
                     delay_loop 2   #服务轮询的时间间隔2s
                    lb_algo rr
                    lb_kind DR
                    protocol TCP
                    real_server 192.168.7.125 80 {
                        weight 1
                        HTTP_GET {
                        url {
                            path /
                            status_code 200
                        }
                        connect_timeout 2
                        nb_get_retry 3
                        delay_before_retry 3
                        }
                    }
                    real_server 192.168.7.126 80 {
                        weight 1
                        HTTP_GET {
                            url {
                                path /
                                status_code 200
                            }
                            connect_timeout 2
                            nb_get_retry 3
                            delay_before_retry 3
                            }
                 }

将上述配置录入/etc/keepalived/keepalived.conf中,过程如下:

[root@DR1 ~]# cd /etc/keepalived/
[root@DR1 keepalived]# cp keepalived.conf{,.bak} #备份下配置文件
[root@DR1 keepalived]# ls
keepalived.conf  keepalived.conf.bak
[root@DR1 keepalived]# vim keepalived.conf
  • 配置中调用的notify脚本内容如下:

    #!/bin/bash
    #
    contact='root@localhost'

    notify() {
        local mailsubject="$(hostname) to be $1, vip floating"
        local mailbody="$(date +'%F %T'): vrrp transition, $(hostname) changed to be $1"
        echo "$mailbody" | mail -s "$mailsubject" $contact
    }

    case $1 in
    master)
        notify master
        ;;
    backup)
        notify backup
        ;;
    fault)
        notify fault
        ;;
    *)
        echo "Usage: $(basename $0) {master|backup|fault}"
        exit 1
        ;;
    esac            
  • chk_down.sh只有条件判断:[[ -f /etc/keepalived/down ]] && exit 1 || exit 0

  1. 配置DR2

DR2,仅需修改如下配置:

  • router_id keepalivedR2

  • state BACKUP

  • priority 90

(2)两台RS配置

两台RS均yum安装nginx,启动服务监听80端口即可。

需要注意的有如下几点:

  • 检查系统是否自带httpd服务并停止,否则会与nginx发生冲突。

  • 添加vip到网卡辅助接口,并修改arp参数抑制apr报文响应,此处可用脚本如下:

    #!/bin/bash
    vip=192.168.7.120
    /usr/sbin/ip addr add $vip/32 dev lo label lo:0
    echo "1" >/proc/sys/net/ipv4/conf/lo/arp_ignore
    echo "2" >/proc/sys/net/ipv4/conf/lo/arp_announce
    echo "1" >/proc/sys/net/ipv4/conf/all/arp_ignore
    echo "2" >/proc/sys/net/ipv4/conf/all/arp_announce
    

    注意:因客户端在同网段,添加完RS抑制arp响应后,需要将客户端的arp缓存清空防止干扰效果。

  • 修改默认主页,主要是做标记区分RS1/RS2。可采用如下方法:

编辑默认index.html 
在vim末行模式输入:%s/nginx/server1/g回车即可

五、测试实验效果

  1. 测试VIP漂移

    • 先在DR2初始BACKUP开启keepalived观察是否有状态变更:

      [root@DR2 ~]# systemctl start keepalived.service 
      [root@DR2 ~]# systemctl status keepalived.service 
      ?.keepalived.service - LVS and VRRP High Availability Monitor
         Loaded: loaded (/usr/lib/systemd/system/keepalived.service; disabled; vendor preset: disabled)
         Active: active (running) since Sun 2018-07-15 13:28:10 CST; 19s ago
        Process: 2620 ExecStart=/usr/sbin/keepalived $KEEPALIVED_OPTIONS (code=exited, status=0/SUCCESS)
       Main PID: 2621 (keepalived)
         CGroup: /system.slice/keepalived.service
                 ?..2621 /usr/sbin/keepalived -D
                 ?..2622 /usr/sbin/keepalived -D
                 ?..2623 /usr/sbin/keepalived -D
      
      Jul 15 13:28:14 DR2 Keepalived_vrrp[2623]: Sending gratuitous ARP on ens39 for 192.168.7.120
      Jul 15 13:28:14 DR2 Keepalived_vrrp[2623]: Sending gratuitous ARP on ens39 for 192.168.7.120
      Jul 15 13:28:14 DR2 Keepalived_vrrp[2623]: Sending gratuitous ARP on ens39 for 192.168.7.120
      Jul 15 13:28:14 DR2 Keepalived_vrrp[2623]: Opening script file /etc/keepalived/notify.sh
      Jul 15 13:28:19 DR2 Keepalived_vrrp[2623]: Sending gratuitous ARP on ens39 for 192.168.7.120
      Jul 15 13:28:19 DR2 Keepalived_vrrp[2623]: VRRP_Instance(VI_1) Sending/queueing gratuitous ARPs on ens39 for 192.168.7.120
      [root@DR2 ~]# mail
      Heirloom Mail version 12.5 7/5/10.  Type ? for help.
      "/var/spool/mail/root": 6 messages 2 new 2 unread
      >N  5 root                  Sun Jul 15 13:28  18/667   "DR2 to be backup, vip floating"
       N  6 root                  Sun Jul 15 13:28  18/667   "DR2 to be master, vip floating"
      & 6
      Message  6:
      From root@DR2.localdomain  Sun Jul 15 13:28:15 2018
      Return-Path: <root@DR2.localdomain>
      X-Original-To: root@localhost
      Delivered-To: root@localhost.localdomain
      Date: Sun, 15 Jul 2018 13:28:14 +0800
      To: root@localhost.localdomain
      Subject: DR2 to be master, vip floating
      User-Agent: Heirloom mailx 12.5 7/5/10
      Content-Type: text/plain; charset=us-ascii
      From: root@DR2.localdomain (root)
      Status: R
      
      2018-07-15 13:28:14: vrrp transition, DR2 changed to be master

      可以看到,因为master没上线,DR2由backup变为master。我们的notify.sh脚本成功发送状态变更通知到本地邮箱。

    • 测试DR1上线效果:

      [root@DR2 ~]# systemctl status keepalived.service 
      Jul 15 14:52:48 DR2 Keepalived_vrrp[1812]: VRRP_Instance(VI_1) Received advert with higher priority 99, ours 90
      Jul 15 14:52:48 DR2 Keepalived_vrrp[1812]: VRRP_Instance(VI_1) Entering BACKUP STATE
      Jul 15 14:52:48 DR2 Keepalived_vrrp[1812]: VRRP_Instance(VI_1) removing protocol VIPs.
      Jul 15 14:52:48 DR2 Keepalived_vrrp[1812]: Opening script file /etc/keepalived/notify.sh
      You have new mail in /var/spool/mail/root
      [root@DR2 ~]# ip addr show
      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: ens39: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
          link/ether 00:50:56:3b:a3:7c brd ff:ff:ff:ff:ff:ff
          inet 192.168.7.122/24 brd 192.168.7.255 scope global ens39
             valid_lft forever preferred_lft forever
          inet6 fe80::250:56ff:fe3b:a37c/64 scope link 
             valid_lft forever preferred_lft forever
      [root@DR1 ~]# systemctl status keepalived.service
      Jul 15 14:52:49 DR1 Keepalived_vrrp[3471]: Sending gratuitous ARP on ens39 for 192.168.7.120
      Jul 15 14:52:49 DR1 Keepalived_vrrp[3471]: Opening script file /etc/keepalived/notify.sh
      Jul 15 14:52:54 DR1 Keepalived_vrrp[3471]: Sending gratuitous ARP on ens39 for 192.168.7.120
      Jul 15 14:52:54 DR1 Keepalived_vrrp[3471]: VRRP_Instance(VI_1) Sending/queueing gratuitous ARPs on ens39 for 192.168.7.120

      DR2因为收到了更高优先级的VRRP组播报文而卸载VIP,DR1获取到VIP。

    • 手工touch生成/etc/keepalived/down文件模拟单机故障也会发生上述VIP漂移现象。

      Jul 15 15:21:51 DR1 Keepalived_vrrp[3907]: WARNING - default user 'keepalived_script' for script execution does not exist - please create.
      # 此处踩坑了,提示conf文件中配置的脚本需要keepalived_script用户执行
      # 还有个巨坑是参考资料中直接在keepalived.conf中定义script "[[ -f /etc/keepalived/down ]] && # exit 1 || exit 0" 但是我多次测试不成功,将条件判断放入脚本,conf文件仅引用路径才成功......
      Jul 15 15:54:24 DR1 Keepalived_vrrp[4438]: /etc/keepalived/chk_down.sh exited with status 1
      Jul 15 15:54:24 DR1 Keepalived_vrrp[4438]: VRRP_Script(chk_down) failed
      Jul 15 15:54:25 DR1 Keepalived_vrrp[4438]: VRRP_Instance(VI_1) Changing effective priority from 99 to 84
      Jul 15 15:54:25 DR1 Keepalived_vrrp[4438]: /etc/keepalived/chk_down.sh exited with status 1
      Jul 15 15:54:25 DR1 Keepalived_vrrp[4438]: VRRP_Instance(VI_1) Received advert with higher priority 90, ours 84
      Jul 15 15:54:25 DR1 Keepalived_vrrp[4438]: VRRP_Instance(VI_1) Entering BACKUP STATE
      Jul 15 15:54:25 DR1 Keepalived_vrrp[4438]: VRRP_Instance(VI_1) removing protocol VIPs.
      Jul 15 15:54:25 DR1 Keepalived_vrrp[4438]: Opening script file /etc/keepalived/notify.sh
  2. 测试RS故障检测

    • 先测试RS是否正常被轮询 for i in {1,5};do curl http://192.168.7.120 ;done

    • 手动关停一台rs的nginx

      [root@RS2 ~]# systemctl stop nginx
      [root@DR1 ~]# journalctl -xe
      Jul 15 16:59:04 DR1 Keepalived_healthcheckers[4437]: Error connecting server [192.168.7.126]:80.
      [root@DR1 ~]# systemctl status  keepalived
      Jul 15 16:59:05 DR1 Keepalived_healthcheckers[4437]: Error connecting server [192.168.7.126]:80.
      Jul 15 16:59:06 DR1 Keepalived_healthcheckers[4437]: Error connecting server [192.168.7.126]:80.
      Jul 15 16:59:07 DR1 Keepalived_healthcheckers[4437]: Error connecting server [192.168.7.126]:80.
      Jul 15 16:59:07 DR1 Keepalived_healthcheckers[4437]: Check on service [192.168.7.126]:80 failed after 3 retry.
      Jul 15 16:59:07 DR1 Keepalived_healthcheckers[4437]: Removing service [192.168.7.126]:80 from VS [192.168.7.120]:80
      Jul 15 16:59:07 DR1 Keepalived_healthcheckers[4437]: Remote SMTP server [127.0.0.1]:25 connected.
      Jul 15 16:59:07 DR1 Keepalived_healthcheckers[4437]: SMTP alert successfully sent.
      [root@DR1 ~]# mail
      N 16 keepalived@localhost  Sun Jul 15 16:59  17/646   "[keepalivedR1] Realserver [192.168.7.126]:80 - DOWN"
      • 重新启动nginx

        Jul 15 17:02:55 DR1 Keepalived_healthcheckers[4437]: HTTP status code success to [192.168.7.126]:80 url(1).
        Jul 15 17:02:55 DR1 Keepalived_healthcheckers[4437]: Remote Web server [192.168.7.126]:80 succeed on service.
        Jul 15 17:02:55 DR1 Keepalived_healthcheckers[4437]: Adding service [192.168.7.126]:80 to VS [192.168.7.120]:80
        Jul 15 17:02:55 DR1 Keepalived_healthcheckers[4437]: Remote SMTP server [127.0.0.1]:25 connected.
        Jul 15 17:02:55 DR1 Keepalived_healthcheckers[4437]: SMTP alert successfully sent.
        You have new mail in /var/spool/mail/root

        重新上线的RS可以顺利被检测到并上线服务,至此,本次实验结束。

 

 

本文来自投稿,不代表Linux运维部落立场,如若转载,请注明出处:http://www.178linux.com/103020

(1)
net31_行者无疆net31_行者无疆
上一篇 2018-07-15
下一篇 2018-07-15

相关推荐