Issue with external Zabbix active hosts - nodata trigger not working

[lucalex]

Hello peps,


I have setup a zabbix server in my internal network and set active agents on servers in external networks (reachable via Public IP and internal NAT port forwarding).
Everything is working fine, agents showing online and items values received.


However, I've noticed that if I have more than 1 server under the same external network, I'm not receiving a nodata/unreachable host trigger if one server is down, it always shows as available, ICM ping is always successful as it pings the public IP.
This problem is not occurring if I have a proxy zabbix server on that specific network.

Has anyone come across this issue before?


Thanks everyone for your help!

[ludama]

I do not know if my problem could in some way be related to your...

I am running a zabbix server 5.4 from official docker compose reachable via Public IP and internal NAT port forwarding.
If i send data to trapper items with zabbix sender from internal network everything is working fine but if try sending data with the hostname i do not receive server response.

The web server in port 80 is working without problem from outside but port 10051 does not seem reachable even with the same router and settings.
Previous version of zabbix docker compose where working fine for both service with this router/nat setup
I hope this information can help in some way