Nagios Support Forum

Hi,

I am only interested in Federated Monitoring that it seems to me a project carried out by people outside nagios and not very easy to maintain.
Is the project valid? Do you recommend it?

I would like to monitor systems behind very unstable VPNs.
I would like to insert a nagios and send alarms to the central nagios.

If connectivity fails, how does it behave? Do I lose all the history?

Will there be a native solution to the nagios proxy core with nagiosXI 6? A thing similar to Zabbix?

Thanks
Federico

Hi,

Sorry,I would like to understand.

So is the distributed monitoring functionality natively implemented with the new version of Nagios6?

Best regards,

Federico

Hi,

Thanks for contacting the support team.

Regarding your setup, have you considered using passive checks? You can install NCPA on those VM's and configure it to send checks passively.

See: Using NCPA For Passive Checks

At this point, we are not planning a native distributed monitoring feature in Nagios XI 6. That said, we are making changes under the hood to allow for more scalability in the future.

Let me know if you have more questions about passive checks or distributed monitoring.

--Benjamin

Hi Benjamin,

thanks for answering.

I think it is a serious major problem for an Enterprise product such as NagiosXI.
Other software eg Zabbix, Icinga, check_mk etc already have this native implementation. (I wouldn't be comparing the software, I like Nagios)

It is not for the number of hosts / services monitored, but for the diversity of the network.
We monitor a lot via VPN and setting up dependencies or parents many times is complicated by the nature of the networks.
Then there are a lot of alarms when a vpn goes down we lose the history of the machines and we have no visibility of things.

Configuring passive checks are not exactly the same as having an "agent" on site that collects everyone's metrics and sends them to the master.

Having said that, given the cost of the license for the new year and the few news / advertisements of the software, my executives are pushing a lot to find other solutions towards a distributed "federated monitoring".

The mod-gearman solution looks nice, but there is little documentation and in the end it is an external product that you don't even want to support.
For an enterprise, such a thing is impossible to put into production.

But why didn't you think about implementing such a thing?

Regards,
-Federico

Hi Federico,

Thank you for choosing Nagios, we do really appreciate it. And we are planning to move towards a more scalable, architecture but that will be after the 6.0 release. For that release, we are completing re-writing the performance graph processor, bringing in a new time-series database, and creating a new visual experience using modern components (React).

Additionally, we working on the underlying API in 6.0 to allow us to break up the application into smaller components for more scalability. I would keep an eye on the roadmaps page for more detailed updates to those plans.

https://www.nagios.com/roadmaps/

Configuring passive checks are not exactly the same as having an "agent" on site that collects everyone's metrics and sends them to the master.

There's another option here and that would be the use of Inbound/Outbound transfers. Let me know if you think that would work in your environment.

Nagios – Monitoring Architecture Solutions For Large Organizations

Configuring Inbound Checks

Configuring Outbound Checks

--Benjamin