docker部署zabbix 6.0高可用集群实验

0 实验环境

虚拟机，postgresql本地部署，zabbix server及nginx容器部署

1 postgresql

参看前作 《postgresql + timescaledb离线安装笔记》完成部署，对外端口tcp 5432，账号zabbix，密码123

2 zabbix server

2.1 拉取镜像

docker pull zabbix/zabbix-server-pgsql:6.0-alpine-latest

2.2 创建网络

docker network create --subnet 172.20.0.0/16 --ip-range 172.20.240.0/20 zabbix-net

2.3 启动server容器

启动两个容器分别作为主备server节点，分别使用本地tcp 10051和10052端口

2.3.1 server主节点

docker run --name zabbix-server-pgsql-1 -t \
	-e DB_SERVER_HOST="172.17.0.1" -e DB_SERVER_PORT="5432" \
	-e POSTGRES_USER="zabbix" -e POSTGRES_PASSWORD='123' -e POSTGRES_DB="zabbix" \
	-e ZBX_CACHESIZE="128M" -e ZBX_HISTORYCACHESIZE="32M" -e ZBX_HISTORYINDEXCACHESIZE="8M" -e ZBX_TRENDCACHESIZE="8M" -e ZBX_VALUECACHESIZE="64M" \
	-e ZBX_LOGSLOWQUERIES="3000" -e ZBX_STARTPOLLERS="5" -e ZBX_STARTPREPROCESSORS="10" -e ZBX_STARTPOLLERSUNREACHABLE="5" -e ZBX_STARTESCALATORS="5" -e ZBX_STARTDBSYNCERS="5" \
	-e ZBX_HANODENAME="server-01" -e ZBX_NODEADDRESS="172.20.240.1" \
	-p 10051:10051 --network=zabbix-net --restart unless-stopped \
	-d zabbix/zabbix-server-pgsql:6.0-alpine-latest

2.3.2 server备节点

docker run --name zabbix-server-pgsql-2 -t \
	-e DB_SERVER_HOST="172.17.0.1" -e DB_SERVER_PORT="5432" \
	-e POSTGRES_USER="zabbix" -e POSTGRES_PASSWORD='123' -e POSTGRES_DB="zabbix" \
	-e ZBX_CACHESIZE="128M" -e ZBX_HISTORYCACHESIZE="32M" -e ZBX_HISTORYINDEXCACHESIZE="8M" -e ZBX_TRENDCACHESIZE="8M" -e ZBX_VALUECACHESIZE="64M" \
	-e ZBX_LOGSLOWQUERIES="3000" -e ZBX_STARTPOLLERS="5" -e ZBX_STARTPREPROCESSORS="10" -e ZBX_STARTPOLLERSUNREACHABLE="5" -e ZBX_STARTESCALATORS="5" -e ZBX_STARTDBSYNCERS="5" \
	-e ZBX_HANODENAME="server-02" -e ZBX_NODEADDRESS="172.20.240.2" \
	-p 10052:10051 --network=zabbix-net --restart unless-stopped \
	-d zabbix/zabbix-server-pgsql:6.0-alpine-latest

2.3.3 调试命令(进入active容器，即zabbix-server-pgsql-1)

显示Server集群状态：

zabbix_server -R ha_status

结果如下

Failover delay: 60 seconds
Cluster status:
   #  ID                        Name                      Address                        Status      Last Access
   1. clkc8ouam00016nrvjcuugxsr server-01                 172.20.240.1:10051             active      4s
   2. clkc8ozws00016nmr5dyv7qd4 server-02                 172.20.240.2:10051             standby     2s

可删除备节点

zabbix_server -R ha_remove_node=clkc8ozws00016nmr5dyv7qd4

3 web

3.1 拉取镜像

docker pull zabbix/zabbix-web-nginx-pgsql:6.0-alpine-latest

3.1 启动容器

注意ZBX_SERVER_HOST和ZBX_SERVER_PORT一定要设置为空

docker run --name zabbix-web-nginx-pgsql -t \
    -e ZBX_SERVER_HOST="" -e ZBX_SERVER_PORT="" \
    -e DB_SERVER_HOST="172.17.0.1" -e DB_SERVER_PORT="5432" \
    -e POSTGRES_USER="zabbix" -e POSTGRES_PASSWORD="123" -e POSTGRES_DB="zabbix" \
    -e ZBX_SERVER_NAME="zabbix-test" \
    -e PHP_TZ="Asia/Shanghai" \
    -p 8080:8080 --network=zabbix-net --restart unless-stopped \
    -d zabbix/zabbix-web-nginx-pgsql:6.0-alpine-latest

4 测试

4.1 原始状态

在System information页面下可以看到，此时server-01是主用节点

4.2 切换

将zabbix-server-pgsql-1容器暂停

docker stop zabbix-server-pgsql-1

此时可以看到server-01变为stopped状态，server-02成为Active状态

再次启动zabbix-server-pgsql-1容器，则server-01变回Standby状态

5 其他笔记

5.1 zabbix-agent2监控docker

如果出现以下报错cannot fetch data: Get http://1.28/info: dial unix /var/run/docker.sock: connect: permission denied，可能是因为zabbix账号对/var/run/docker.sock不具备权限，我的修复命令如下：

chgrp docker /var/run/docker.sock
usermod -aG docker zabbix
systemctl restart zabbix-agent2