I just need someone to explain the difference between the Icinga2 Master, Client, and Agent. Does the Icinga2 agent need to be installed on both the master and client, or separately from both? I’m sorry, just between the terms host, client, master, node, endpoint, agent, etc. …I’m getting really confused as to how the standard setup should be configured and what is what. Thanks for the help in advance.
I just need someone to explain the difference between the Icinga2 Master,
Client, and Agent.
Firstly, there is no such thing as “the Icinga2 Agent”.
Does the Icinga2 agent need to be installed on both the master and client,
or separately from both?
Where have you come across any reference to “Icinga2 Agent”?
I’m sorry, just between the terms host, client, master, node, endpoint,
agent, etc. …I’m getting really confused as to how the standard setup
should be configured and what is what.
Host = a machine in your network (standard definition, nothign special to
Icinga)
Client = a machine you are monitoring with Icinga2, which reports back to a
Master Icinga2 server
Master = a machine which has Icinga2 installed on it (and probably Icingaweb2
as well), and which communicates with one or more Client machines
Node = something being monitored which may or may not have Icinga2 installed
on it (it might be a switch, being monitored by SNMP, for example)
Endpoint = the hostname or IP address of something in your network
Agent = this terminology is not used by icinga, to the best of my knowledge.
If you only have one server, running icinga, and monitoring itself, then it’s
just a single standalone server, not really either a Master or a Client.
If you have a Master managing one or more Client machines, then they all have
exactly the same “Icinga2” software installed on them (assuming they’re all
Linux machines, of course), and it is simply the configuration files which define
which one is the Master and which are the Clients.
Masters can communicate to Clients in order to send monitoring commands and
get back the results, or Clients can communicate to Masters in order to find
out what monitoring commands to run and return the results. The choice is
yours and generally depends on what network routing is most convenient for you
(especially if you have any NAT involved).
Let us know if you have further questions 
Antony.
1 Like
First of all, thank you so much for taking the time to respond. I got the icinga2 agent reference from here: https://galaxy.ansible.com/search?deprecated=false&keywords=icinga2&order_by=-relevance&page=1&page_size=10 …I’m seeking to use Ansible to automate the setting up of icinga2 and just wanted to make sure I had a clear understanding of the terminology. The icinga2 agent seems to be a separate Ansible script in each implementation. Any info on this would be very, very helpful. Thanks again!
I suggest you ask on an Ansible list or forum, or else email the author of the
script.
Antony.
1 Like
Ok, will do. Thanks again for the help.
Since 2.11, this term is used throughout the docs, to avoid confusion with web interfaces and the client terminology (client server vs master satellite agent).
https://icinga.com/docs/icinga2/latest/doc/16-upgrading-icinga-2/#documentation
Ah, thanks for the clarification.
Since I’ve been unable to upgrade to 2.11 so far (as discussed in other recent
postings), I wasn’t aware that “agent” was now being used in the
documentation.
I thought it had been specifically avoided, partly to remove any potential
confusion with the “Zabbix Agent” which is a small-footprint monitoring agent
used in the Zabbix environment (and distinct from the full Zabbix monitoring
application), whereas Icinga uses the same full software on every machine in
the network.
Antony.