What is the best way to implement monitoring of the availability of several servers for each other?

Question

linux

given: there are 10 servers in different countries and 1 server with a monitoring system, no matter what, nagios, cacti, etc.
task: to organize monitoring of the availability of all servers within this network. so that server "1" polls all 10 servers from its host and says which ones are not available for it, and so on all servers.
I imagine the implementation of this task as follows:
commands are sent from the monitoring server to the servers on some open port to some daemon, the command parameters indicate which hosts need to be polled (query-1-2-3-4...), server "1" queries all 10 servers that were listed in the parameters (query-1-2-3-4...) for availability and returns the result to the monitoring server in the form of the response time of each server (respon-1.32-1.35-1.11-0.245.. .)
the monitoring server receives results from 10 servers and processes
the polling interval = every minute
protocol = TCP
polled port = HTTP which is open on all servers
, so the question is how best to implement this task, is there ready-made software.

Answer 1

Perhaps the answer is in the question:

...no matter what, nagios...

zabbix is ​​still there
which servers to query, and there are more than 200 servers and they change every day, well, or once a week anyway, you can’t overwrite configs like that transfer part of the condition to the text of the question
. Even if you are completely too lazy and set up a gui for nagios, then there is a wonderful clone button, chose whom to clone from, registered a new name and what services you need to monitor, clicked in the browser with the mouse, done.

Answer 2

Simple enough. Install collectd on all servers. It has a plugin https://collectd.org/wiki/index.php/Plugin:Ping
In the plugin configuration, set pings to all your servers.
You give the result to nagios via, for example , https://collectd.org/documentation/manpages/collec... (I haven't tried it!!!)
You can also pour the result into mongo, redis, and rabbitmq. Well, you can get nagios plugins from everywhere, since they are written in 5 minutes.
Or more radically, put graphite (we had it) or unfluxdb + grafana (now) and pour data from collecd servers there.
PS. As a bonus of installing collectd, you will get a fast and resource-free collection of almost any statistics.

Answer 3

Icinga ;) https://www.icinga.org if I can help )
well, you can also take ansible, which will throw scripts on the servers, etc.