Nagios XI Hardware Recomendations

[mpedros]

Dear Support:

I will make an implementation of Nagios XI in a client, I need to monitor 3800 host, the total services to monitor is 8400. Can I install in a VM?

Regrads,

[eloyd]

Yes. Nagios XI will run quite nicely with up to about 5,000-10,000 hosts and between 20,000-50,000 services. However, you will want to give your virtual machine at least 4 cores and a minimum of 8GB of memory plus enough disk space to hold all your logs (we'd recommend 250GB minimum for the configuration you are looking for). If you have to choose between adding more CPU or more memory, always add more memory.

[npolovenko]

In addition to @eloyd recommendations I'd highly suggest to implement a ramdisk. It's fairly easy to implement but it really boosts the performance of larger systems.
https://assets.nagios.com/downloads/nag ... giosXI.pdf

Also, this article will walk you through some tweaks to maximize your XI performance:
https://assets.nagios.com/downloads/nag ... 2hEALw_wcB

[mpedros]

Yes. Nagios XI will run quite nicely with up to about 5,000-10,000 hosts and between 20,000-50,000 services. However, you will want to give your virtual machine at least 4 cores and a minimum of 8GB of memory plus enough disk space to hold all your logs (we'd recommend 250GB minimum for the configuration you are looking for). If you have to choose between adding more CPU or more memory, always add more memory.

Dear @eloyd Thank you very much for your help, it is very useful for me. Regards!

[mpedros]

In addition to @eloyd recommendations I'd highly suggest to implement a ramdisk. It's fairly easy to implement but it really boosts the performance of larger systems.
https://assets.nagios.com/downloads/nag ... giosXI.pdf

Also, this article will walk you through some tweaks to maximize your XI performance:
https://assets.nagios.com/downloads/nag ... 2hEALw_wcB

Dear @npolovenko, thank you very much for your suggestion. I'm already reading the documents. Regards!

[npolovenko]

@mpedros, You're welcome. Keep in mind that the estimate numbers that eloyd gave you are dependent on check times, your hardware and specs (not all CPUs and RAM are created equally), and the types of checks (compiled vs. interpreted and active vs. passive). Let us know if have any other questions so far.

[eloyd]

Yes, remember, I said "Nagios XI will run quite nicely" but that doesn't mean that it will! Only YOU can determine if your machine is up to the task by checking the specs of its resources.

[npolovenko]

@mpedros, Do you have any other questions for us so far?

[mpedros]

Dear @eloyd and @npolovenko, thank you very much for your help, for the moment I have no more questions. I will monitor the hardware as we add the nodes. Regards.

[eloyd]

Here's something to help you do that: Inside the Nagios XI Configuration Wizard (Config -> Configuration Wizards) is a wizard called "Nagiostats Wizard." RUN THIS ONE! It's basically Nagios monitoring itself as a monitoring tool, as opposed to Nagios monitoring its own operating system.

Specifically, I recommend that you monitor the following:

• Average Active Host Latency
• Average Active Service Latency
• Average Host Execution Time
• Average Service Execution Time
• Active Host Checks 15min
• Active Service Checks 15min

The first two should cause alarm if they are greater than 1000ms (1 second). This means that Nagios is having problems getting host and service checks to execute.

The second two should cause alarm if they are greater than maybe 20,000ms (20 seconds). You will want to adjust this number as you add services and hosts. This indicates, on average, how long it takes to complete a host or service check.

The last two are mainly just counters that you can see over time to get an idea of how many host and service checks you are performing averaged over 15 minute intervals.

All of these data can help you with capacity planning as far as resources to add to the machine.

Good luck!