Nagios Support Forum

Hi,

Thanks for your advices, but I think there's one final point that I misunderstand (see below).

Your last post about the distributed architecture, you have to think of the checks being run on the remote gearman worker and the IP address that have to be checked have to be accessible from that system.

I aggree, but how do I configure these hosts on the central server in order to be able to supervise them in Nagios.

When configuring a host, Nagios want at least a host name and an address. If a remote host is supervised by a distant worker, in my architecture it will surely mean that it's not accesible from the central server (behind a router and/or a firewall, in another LAN with a private IP address).
So, what would my host configuration look like in Nagios central server ?

Thanks four your responses.

The host entry would look like any other host entry.
You would give it a host name and an IP address like any other host but the IP address would have to be accessible from the remote Gearman Worker.
Then you would add that Host entry to the Hostgroup that is configured to run on the remote Gearman worker.
That way, it will only run on that system.

Ok, I understand. This means that I can have several hosts with a same private IP address in my Nagios hosts' configuration. Just have to take care that they are in different hostgroups, and each hostgroup handled by a different worker.

But this means, as hosts configuration is centralized on the Nagios Server, that workers must get the informations about the hosts (namely IP addresss) from the server.
And what if, like I said when I was talking about distributed architecture, the host on which a worker is running, is unreachable from the Nagios server ? Understand that it's on a private customer network in which I'm not granted to have any access.

I think this is the last point to really understand how mod_gearman works.

Thanks for your responses.

The Nagios server does not need to access the remote network, only the remote worker needs to access the Nagios Gearman server.
What happens is the Gearman Worker, connects to the Gearman Server and retrieves the jobs from the queue and runs them.
The gearman server only queues the jobs on the server, the workers has to retrieve them and run them.

Ok now I think it's completly clear !

Thanks a lot for your help.

Now I'm trying to install a worker on a centOS virtual host. But when issuing the command "yum install mod-gearman-worker", I get a "No available packets" message.

If ever, does somone know on which repository I can find this packet.

Thanks.

Depending on which Yum Repository you have enabled on your system, the Gearman package could be called mod_gearman so run the following on your system to see if it is available.

Thanks for responding...

I have two additional demands :
- Does anybody know the average number of metrics I can monitor on a raspberry (Pi 3 1gb/16gb), without overload issues ?
- Is there somewhere any documentation about the way mode gearman encrypt datas. Is this a light or heavy encryption method ?
I made some researches on the web (including www.mode-gearman.org)...didn't find anything relevant.

Thanks.

Your welcome. I'll leave the post open in case someone has any information on your questions.