keepalived搭建zabbix server双机高可用

头像
Buttonwood
    阅读 8 分钟
    1

    零. 准备工作

    最近公司zabbix服务器升级，鉴于以前单机负载压力较大，宕机之后系统瘫痪，趁机改成双机高可用架构。
    以下是用keepalived做HA的搭建过程，以备参考！

    两台主机，三个ip，两台主机配置zabbix和keepalived用于高可用架设。

    IP地址 主机名 备注
    192.168.3.141 zabbix_master zabbix/keepalived
    192.168.3.141 zabbix_slaver zabbix/keepalived
    192.168.3.144 VIP ---

    一. keepalived安装

    1. Installation

    centOS 7下安装很简单，直接yum即可。

    $ su root 
$ yum install keepalived
$ /usr/sbin/keepalived -D
$ pstree |grep keepalived
        |-keepalived---2*[keepalived]

    可以看到实际启动了三个进程。

    2. Base Configuration

    制作开机启动

    pgrep keepalived |xargs kill -15
 #重新加载
systemctl daemon-reload  
 #设置开机自动启动
systemctl enable keepalived.service
 #取消开机自动启动  
systemctl disable keepalived.service 
 #启动
systemctl start keepalived.service
 #停止
systemctl stop keepalived.service
 #状态检查
systemctl status keepalived.service
    3. Advance Configuration
    su root
chown zabbix:zabbix /usr/sbin/keepalived
chown zabbix:zabbix /lib/systemd/system/keepalived.service
chown zabbix:zabbix /etc/keepalived/keepalived.conf
su zabbix 
vim /lib/systemd/system/keepalived.service
vim /etc/keepalived/keepalived.conf
chmod +x /home/zabbix/keepalived/src/zabbix_status_check2.sh

    主要是keepalived.conf和zabbix_status_check2.sh两个脚本，网上资料很多，这里不细讲。

    二. 双机启动

    1. 主机启动keepalived
    Mar 13 17:27:57 vm1184 Keepalived_vrrp[5026]: VRRP_Script(zabbix_status_check) succeeded
Mar 13 17:28:00 vm1184 Keepalived_vrrp[5026]: VRRP_Instance(VI_1) Transition to MASTER STATE
Mar 13 17:28:01 vm1184 Keepalived_vrrp[5026]: VRRP_Instance(VI_1) Entering MASTER STATE
Mar 13 17:28:01 vm1184 Keepalived_vrrp[5026]: VRRP_Instance(VI_1) setting protocol VIPs.
Mar 13 17:28:01 vm1184 Keepalived_vrrp[5026]: VRRP_Instance(VI_1) Sending gratuitous ARPs on ens160 for 192.168.3.144
Mar 13 17:28:01 vm1184 Keepalived_healthcheckers[5025]: Netlink reflector reports IP 192.168.3.144 added
Mar 13 17:28:01 vm1184 avahi-daemon[637]: Registering new address record for 192.168.3.144 on ens160.IPv4.
Mar 13 17:28:06 vm1184 Keepalived_vrrp[5026]: VRRP_Instance(VI_1) Sending gratuitous ARPs on ens160 for 192.168.3.144
    2. 主机地址,VIP已经绑定到网卡
    $  ip a
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: ens160: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:50:56:94:eb:ad brd ff:ff:ff:ff:ff:ff
    inet 192.168.3.141/32 brd 192.168.3.141 scope global ens160
       valid_lft forever preferred_lft forever
    inet 192.168.3.144/24 scope global ens160
       valid_lft forever preferred_lft forever
    inet6 fe80::ff15:1cc9:5bd0:b06e/64 scope link
       valid_lft forever preferred_lft forever
    3. 备机日志，keepalived已启动。
    $  tail -f  /var/log/messages
Mar 14 09:24:02 vm1185 Keepalived_healthcheckers[7588]: Registering Kernel netlink command channel
Mar 14 09:24:02 vm1185 Keepalived_healthcheckers[7588]: Opening file '/etc/keepalived/keepalived.conf'.
Mar 14 09:24:02 vm1185 Keepalived_healthcheckers[7588]: Configuration is using : 8087 Bytes
Mar 14 09:24:02 vm1185 Keepalived_healthcheckers[7588]: Using LinkWatch kernel netlink reflector...
    4. 备机ip地址，VIP实际并没有绑到网卡上。
    $  ip a
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: ens160: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:50:56:94:88:c9 brd ff:ff:ff:ff:ff:ff
    inet 192.168.3.142/32 brd 192.168.3.142 scope global ens160
       valid_lft forever preferred_lft forever
    inet6 fe80::e00c:f722:cd30:ad8b/64 scope link
       valid_lft forever preferred_lft forever

    三. 主备切换

    1. 杀掉主机上的keepalived之后，备机现在自动切换成主机了。
    Mar 14 09:29:04 vm1185 Keepalived_vrrp[7589]: VRRP_Instance(VI_1) Transition to MASTER STATE
Mar 14 09:29:05 vm1185 Keepalived_vrrp[7589]: VRRP_Instance(VI_1) Entering MASTER STATE
Mar 14 09:29:05 vm1185 Keepalived_vrrp[7589]: VRRP_Instance(VI_1) setting protocol VIPs.
Mar 14 09:29:05 vm1185 Keepalived_vrrp[7589]: VRRP_Instance(VI_1) Sending gratuitous ARPs on ens160 for 192.168.3.144
Mar 14 09:29:05 vm1185 Keepalived_healthcheckers[7588]: Netlink reflector reports IP 192.168.3.144 added
Mar 14 09:29:05 vm1185 avahi-daemon[674]: Registering new address record for 192.168.3.144 on ens160.IPv4.
Mar 14 09:29:10 vm1185 Keepalived_vrrp[7589]: VRRP_Instance(VI_1) Sending gratuitous ARPs on ens160 for 192.168.3.144
    2. 现主机（原备机）现在的网卡情况，VIP已经绑定上去了。
    $  ip a
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: ens160: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:50:56:94:88:c9 brd ff:ff:ff:ff:ff:ff
    inet 192.168.3.142/32 brd 192.168.3.142 scope global ens160
       valid_lft forever preferred_lft forever
    inet 192.168.3.144/24 scope global ens160
       valid_lft forever preferred_lft forever
    inet6 fe80::e00c:f722:cd30:ad8b/64 scope link
       valid_lft forever preferred_lft forever
    3. 再启动原主机上的keepalived，原主机变备机，VIP绑定被清除了。
    $  tail -f /var/log/messages
Mar 14 09:30:12 vm1184 Keepalived_vrrp[11857]: VRRP_Instance(VI_1) Entering MASTER STATE
Mar 14 09:30:12 vm1184 Keepalived_vrrp[11857]: VRRP_Instance(VI_1) setting protocol VIPs.
Mar 14 09:30:12 vm1184 Keepalived_vrrp[11857]: VRRP_Instance(VI_1) Sending gratuitous ARPs on ens160 for 10.18      .8.144
Mar 14 09:30:12 vm1184 avahi-daemon[637]: Registering new address record for 192.168.3.144 on ens160.IPv4.
Mar 14 09:30:12 vm1184 Keepalived_healthcheckers[11856]: Netlink reflector reports IP 192.168.3.144 added
Mar 14 09:30:12 vm1184 Keepalived_vrrp[11857]: VRRP_Instance(VI_1) Received higher prio advert
Mar 14 09:30:12 vm1184 Keepalived_vrrp[11857]: VRRP_Instance(VI_1) Entering BACKUP STATE
Mar 14 09:30:12 vm1184 Keepalived_vrrp[11857]: VRRP_Instance(VI_1) removing protocol VIPs.
Mar 14 09:30:12 vm1184 Keepalived_healthcheckers[11856]: Netlink reflector reports IP 192.168.3.144 removed
Mar 14 09:30:12 vm1184 avahi-daemon[637]: Withdrawing address record for 192.168.3.144 on ens160.
    4. 原主机变备机，VIP绑定确实被清除了。
     $  ip a
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: ens160: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:50:56:94:eb:ad brd ff:ff:ff:ff:ff:ff
    inet 192.168.3.141/32 brd 192.168.3.141 scope global ens160
       valid_lft forever preferred_lft forever
    inet6 fe80::ff15:1cc9:5bd0:b06e/64 scope link
       valid_lft forever preferred_lft forever

    四. zabbix_server 宕机实验

      # 原server启动方式
$  ps -ef |grep zabbix_server

zabbix   31668 31643  0 Mar13 ?        00:00:11 ./sbin/zabbix_server: escalator #1 [processed 0 escalations in 0.001148 sec, idle 3 sec]
zabbix   31669 31643  0 Mar13 ?        00:00:03 ./sbin/zabbix_server: proxy poller #1 [exchanged data with 0 proxies in 0.000066 sec, idle 5 sec]
zabbix   31670 31643  0 Mar13 ?        00:00:09 ./sbin/zabbix_server: self-monitoring [processed data in 0.000046 sec, idle 1 sec]
zabbix   31671 31643  0 Mar13 ?        00:00:03 ./sbin/zabbix_server: task manager [processed 0 task(s) in 0.000289 sec, idle 5 sec]

  # 杀掉进程，操作宕机
$  pgrep zabbix_server |xargs kill -15

  # 检查
$  ps -ef |grep zabbix_server

zabbix   14053 14028  0 09:59 ?        00:00:00 /home/zabbix/zabbix/sbin/zabbix_server: history syncer #4 [synced 0 items in 0.000043 sec, idle 1 sec]
zabbix   14054 14028  0 09:59 ?        00:00:00 /home/zabbix/zabbix/sbin/zabbix_server: escalator #1 [processed 0 escalations in 0.001202 sec, idle 3 sec]
zabbix   14055 14028  0 09:59 ?        00:00:00 /home/zabbix/zabbix/sbin/zabbix_server: proxy poller #1 [exchanged data with 0 proxies in 0.000057 sec, idle 5 sec]
zabbix   14056 14028  0 09:59 ?        00:00:00 /home/zabbix/zabbix/sbin/zabbix_server: self-monitoring [processed data in 0.000040 sec, idle 1 sec]
zabbix   14057 14028  0 09:59 ?        00:00:00 /home/zabbix/zabbix/sbin/zabbix_server: task manager [processed 0 task(s) in 0.000433 sec, idle 5 sec]

    说明keepalived的健康检查脚本确实执行了，检查不到zabbix_server进程的情况下会自动重启。

    zabbixkeepalivedcentos7
    Buttonwood
    17 声望1 粉丝

    Float like a butterfly! Stand like a buttonwood!

    « 上一篇
    自动检测硬盘，格式化及挂载
    下一篇 »
    zabbix3.2和oracle数据库安装记录

    引用和评论

    推荐阅读
    头像
    分区partition知识点

    Buttonwood阅读 2.3k

    头像
    CentOS7 Docker安装Gitlab

    YYGP阅读 6.6k

    头像
    【Docker】基本概念及语法与环境搭建

    neojayway阅读 4.8k

    头像
    Centos防火墙和端口操作

    zxl20070701阅读 1.2k

    头像
    Ubuntu 安装和配置 Zabbix Agent

    鸠摩智首席音效师阅读 376

    头像
    Ubuntu 24.04 上安装和配置 Zabbix 服务器

    鸠摩智首席音效师阅读 356

    头像
    信创实践(3)：基于x2openEuler将CentOS升级成openEuler，享受其带来的创新和安全特性

    AbelEthan阅读 280

    0 条评论
    得票最新