Page 1 of 1

How does Nagios handle hardware upgrades?

Posted: Mon Dec 19, 2011 7:08 pm
by cwscribner
Hi all.

I recently received the specs for my new Nagios server and its pretty awesome. Max of 32 CPUs, 512GB RAM, and countless TB of HDD space. I have two questions in relation to this info:

1) Does Nagios care what sort of hardware its running on? For example, will upgrading from 8 cores to 16 cores kill Nagios or will it just use up the extra power to process checks?

2) Does Nagios have comprehensive resource calculation to determine actual usage? For example, assuming my previous upgrade to 16 cores, will Nagios calculations take into account the multiple cores or does it hit a ceiling that it can't calculate beyond? (Are the calculations published anywhere?)

Re: How does Nagios handle hardware upgrades?

Posted: Tue Dec 20, 2011 12:01 pm
by mguthrie
I had to run this one by our lead developer, I'm a bit out of my league on this question ; )

Re: How does Nagios handle hardware upgrades?

Posted: Wed Dec 21, 2011 4:12 pm
by admin
Yes, Nagios Core can make use of multiple processors/cores.

The main "thread" can't, but that is only used for event scheduling. The actual host and service checks are forked off the main thread and may be run in other processors/cores. Ultimately, its ups to the kernel to decide which proc/core the checks run in.

Upgrading/downgrading processors/cores won't harm/benefit Nagios other than affecting performance as it would with other applications

Hope that helps.

Re: How does Nagios handle hardware upgrades?

Posted: Wed Dec 21, 2011 4:23 pm
by cwscribner
That does help, thanks. My client has decided to upgrade to an AMD Opteron 6274 16-core so I'm glad to know it'll work.

Can you elaborate any further on the calculations that Nagios does to generate usage stats? Is there a point in which hardware will exceed the monitoring capabilities?

Re: How does Nagios handle hardware upgrades?

Posted: Thu Dec 22, 2011 12:32 pm
by mguthrie
Can you elaborate any further on the calculations that Nagios does to generate usage stats?
What info in-particular are you needing?

Code: Select all

Is there a point in which hardware will exceed the monitoring capabilities?
I've met people who have scaled Nagios Core installations to over 1 million services, so it all depends on what you're monitoring and ultimately how many hosts and services you'll eventually need to monitor.