When monitoring remote Zabbix servers the standard OS template will cover unreachability, hardware/OS failures while "Remote Zabbix server health" will alert on performance problems.

It would be be useful if the latter template also included a trigger when no data is received / it can't reach the remote zabbix, so that the case is covered when the OS functioning but e.g. zabbix-server has crashed.