最近公司zabbix服務器升級,鑑於之前單機負載壓力較大,宕機以後系統癱瘓,趁機改爲雙機高可用架構。
如下是用keepalived作HA的搭建過程,以備參考!vim
兩臺主機,三個ip,兩臺主機配置zabbix和keepalived用於高可用架設。服務器
IP地址 | 主機名 | 備註 |
---|---|---|
192.168.3.141 | zabbix_master | zabbix/keepalived |
192.168.3.141 | zabbix_slaver | zabbix/keepalived |
192.168.3.144 | VIP | --- |
centOS 7下安裝很簡單,直接yum
便可。架構
$ su root $ yum install keepalived $ /usr/sbin/keepalived -D $ pstree |grep keepalived |-keepalived---2*[keepalived]
能夠看到實際啓動了三個進程。oop
製做開機啓動ui
pgrep keepalived |xargs kill -15 #從新加載 systemctl daemon-reload #設置開機自動啓動 systemctl enable keepalived.service #取消開機自動啓動 systemctl disable keepalived.service #啓動 systemctl start keepalived.service #中止 systemctl stop keepalived.service #狀態檢查 systemctl status keepalived.service
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兩個腳本,網上資料不少,這裏不細講。scala
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
$ 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
$ 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...
$ 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
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
$ 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
$ 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.
$ 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
# 原server啓動方式 $ 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] # 殺掉進程,操做宕機 $ pgrep zabbix_server |xargs kill -15 # 檢查 $ 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]
說明keepalived的健康檢查腳本確實執行了,檢查不到zabbix_server進程的狀況下會自動重啓。日誌