I have this running on a Raspberry Pi 5:

services:
  db:
    image: postgres:16-alpine
    environment:
      - POSTGRES_USER=zabbix
      - POSTGRES_PASSWORD=zabbix
      - PGDATA=/var/lib/postgresql/data
    volumes:
      - /opt/docker/zabbix7/pgdata/16/data:/var/lib/postgresql/data
    networks:
      - zabbix7
    restart: unless-stopped

  # fping needs setsuid
  # Connect to container as "root" and run: chmod +s /usr/sbin/fping
  server:
    image: zabbix/zabbix-server-pgsql:alpine-7.4-latest
    environment:
      - POSTGRES_USER=zabbix
      - POSTGRES_PASSWORD=zabbix
      - PHP_TZ=Europe/London
      - ZBX_SERVER_NAME=zabbix.domain.com
      - ZBX_NODEADDRESS=zabbix-server:10051
    cap_add:
      - NET_RAW
      - NET_ADMIN
    volumes:
      - /opt/docker/zabbix7/zabbix-server/alertscripts:/usr/lib/zabbix/alertscripts
      - /opt/docker/zabbix7/zabbix-server/externalscripts:/usr/lib/zabbix/externalscripts
      - /opt/docker/zabbix7/zabbix-server/mibs:/usr/lib/zabbix/mibs
      - /opt/docker/zabbix7/zabbix-server/modules:/usr/lib/zabbix/modules
      - /opt/docker/zabbix7/zabbix-server/export:/var/lib/zabbix/export
      - /opt/docker/zabbix7/zabbix-server/snmptraps:/var/lib/zabbix/snmptraps
    ports:
      - 10051:10051
    depends_on:
      - db
    links:
      - "db:postgres-server"
    networks:
      - zabbix7
      - traefik-public
    restart: unless-stopped

  web:
    image: zabbix/zabbix-web-nginx-pgsql:alpine-7.4-latest
    restart: unless-stopped
    environment:
        - PHP_TZ=Europe/London
        - ZBX_SERVER_NAME=zabbix.domain.com
        - ZBX_SERVER_HOST=zabbix-server
    #ports:
    #  - 10080:8080
    #  - 10081:443
    depends_on:
      - server
      - db
    links:
      - "server:zabbix-server"
      - "db:postgres-server"
    networks:
      - zabbix7
      - traefik-public
    labels:
      traefik.enable: "true"
      traefik.http.routers.zbx.rule: Host(`zabbix.domain.com`)
      traefik.http.routers.zbx.entrypoints: https
      traefik.http.routers.zbx.tls: "true"
      traefik.http.routers.zbx.tls.certresolver: le
      traefik.http.services.zbx.loadbalancer.server.port: "8080"

networks:
  traefik-public:
    external: true
  zabbix7:
    attachable: true

I used to. It's really complicated and not big in the self hosted space.

I'd recommend the standard Prometheus, alert manager, and grafana stack instead. That's what I use now, and although yaml sure is something, at least there's tons of people to copy.

I use it at home, as a test environment as I brought it to 2 workplaces. I have a mysql VM and a front end VM.

It's still on my to-do list.

I am using it and tbh didn't have too many issues with it. It runs as a LXC on my Proxmox server.

With that it's a fairly comfortable setup - it does have API access on the proxmox node and therefore automatically discovers all LXCs,even the ones you add after the installation.

For other machines I use a fairly easy bash script to download the agent 2 and then overwrite the config file with the right parameters,but that's just me being lazy - it's not that much work doing it by hand as well.

And for everything else there is always SNMP which is fairly well supported and there are tons of templates nowadays.

Tbh, I had Prometheus/Grafana before and found it to be much more complicated, especially when you need active and passive nodes. The fact that Zabbix is "All in one" is fairly nice sometimes.

Dashboards are a bit lacking behind Grafana at times,but I can live with that.

I run mine on a separate server in containers (using Podman). For architecture, the server just connects directly to each agent, using my VPN network

have it on dietpi (pi 4) + tailscale at home to monitor my dedics