CentOS Keepalived配置

    借助 Keepalived 完成VIP配置

    • 在网关节点安装 Keepalived
    1. $ yum install -y keepalived

    • 编辑配置文件

      注意！当前调用健康监测脚本内容为注释状态，原因是在安装 Rainbond 前 需要确保 VIP 已经存在；在 Rainbond 安装完成之后需将注释取消，才能实现健康监测，确保 网关高可用。

      • 主节点
    1. $ vi /etc/keepalived/keepalived.conf
    3. ! Configuration File for keepalived
    5. global_defs {
    6.    router_id LVS_DEVEL
    7. }
    8. #vrrp_script check_gateway {
    9. # 检测脚本
    10. #   script "/etc/keepalived/check_gateway_status.sh"
    11. # 执行间隔时间
    12. #   interval 5
    13. #}
    14. vrrp_instance VI_1 {
    16.     #因使用非抢占模式，这里都为backup
    17.     state BACKUP 
    18.     #网卡设备名，通过 ifconfig 命令确定  
    19.     interface ens6f0       
    20.     virtual_router_id 51
    21.     #优先级，主节点大于备节点     
    22.     priority 100    
    23.     advert_int 1
    24.     #非抢占模式
    25.     nopreempt
    26.     authentication {
    27.         auth_type PASS
    28.         auth_pass 1111
    29.     }
    30.     virtual_ipaddress {
    31.         <VIP>                
    32.      }
    33. #     track_script {
    34. #    check_gateway
    35. #   }
    36. }
    • 从节点
    1. $ vi /etc/keepalived/keepalived.conf
    3. ! Configuration File for keepalived
    5. global_defs {
    6.    router_id LVS_DEVEL
    7. }
    8. #vrrp_script check_gateway {
    9. # 检测脚本
    10. #  script "/etc/keepalived/check_gateway_status.sh"
    11. # 执行间隔时间
    12. #  interval 5
    13. #  }    
    14.     vrrp_instance VI_1 {
    15.     #因使用非抢占模式，这里都为backup
    16.     state BACKUP 
    17.     #网卡设备名，通过 ifconfig 命令确定   
    18.     interface ens6f0
    19.     virtual_router_id 51
    20.     #优先级，主节点大于备节点   
    21.     priority 50
    22.     advert_int 1
    23.     #非抢占模式
    24.     nopreempt
    25.     authentication {
    26.         auth_type PASS
    27.         auth_pass 1111
    28.     }
    29.     virtual_ipaddress {
    30.         <VIP>            
    31.     }
    32. #     track_script {
    33. #    check_gateway
    34. #   }
    35. }
    • 健康监测脚本

    扩展对网关节点健康检查的脚本，脚本的功能是当 rbd-gateway 组件停止服务，则关闭本机的 Keepalived，切换 VIP 。(主从都需操作)

    1. $ vi /etc/keepalived/check_gateway_status.sh 
    3. #!/bin/bash                                                                                             
    4. /usr/bin/curl -I http://localhost:10254/healthz 
    6. if [ $? -ne 0 ];then
    8.      cat /var/run/keepalived.pid | xargs kill
    10. fi

    添加执行权限

    1. $ chmod +x /etc/keepalived/check_gateway_status.sh

    • 更改 Keepalived systemd 配置文件，添加两项配置

      添加两项参数

    1. $  vi /lib/systemd/system/keepalived.service
    3. # /lib/systemd/system/keepalived.service
    4. [Unit]
    5. Description=Keepalive Daemon (LVS and VRRP)
    6. After=network-online.target
    7. Wants=network-online.target
    8. # Only start if there is a configuration file
    9. ConditionFileNotEmpty=/etc/keepalived/keepalived.conf
    10. [Service]
    11. Type=forking
    12. KillMode=process
    13. PIDFile=/var/run/keepalived.pid
    14. # Read configuration variable file if it is present
    15. EnvironmentFile=-/etc/default/keepalived
    16. ExecStart=/usr/sbin/keepalived $DAEMON_ARGS
    17. ExecReload=/bin/kill -HUP $MAINPID
    18. # 总是重启该服务
    19. Restart=always
    20. # 重启间隔时间
    21. RestartSec=10
    23. [Install]
    24. WantedBy=multi-user.target

    • 启动服务

      启动服务并设置开机自启动

    1. systemctl start keepalived
    2. systemctl enable keepalived
    3. systemctl status keepalived
    • 检测 VIP 是否启动
    1. ip a |grep <VIP>

    在进行下一步操作前请确保 VIP 已经存在，如果 VIP 不存在，请重新审查本节操作；在 Rainbond 安装完成之后请将配置文件中注释取消并重启 Keepalived 服务，实现健康监测，以此确保 网关高可用。