Nagios XI Load recommendation

[JakeHatMacys]

Hey guys, just a quick question... I tried searching around a bit and didn't seem to find a solid answer. But we're looking at putting up Nagios monitoring for an additional 65,000 devices. I'd say probably 2 to 3 checks per device. What's the current standard on server loads from a software standpoint?

Our VMs are pretty beefy so I don't think hardware will be an issue but we just want to get a feel for how many servers we're going to need to monitor that man devices.

Thoughts?

[rkennedy]

The general recommendation is 32GB of ram with 4-8 CPU cores. This should be able to sustain 20-30k checks, whether it be 10k hosts + 20k services, or however you'd like to aggregate them. This of course would be with some optimization done, such as an offloaded DB, a ram disk, or gearman. (see this link for more of an explanation - https://support.nagios.com/kb/article.php?id=503)

The main limitation gets down to the reporting, where SQL just can't keep up with processing a lot of data.

[JakeHatMacys]

Thanks for the reply, I'll let my boss know and we'll be contacting sales for more licenses soon O.o

[dwhitfield]

There's more issues than just MySQL, but https://www.mysql.com/products/enterprise/database/ is something to look forward to in the future.

Did you need any help with any of the performance suggestions? If not, are we ready to close this thread?

[JakeHatMacys]

Sure you can close, appreciate it.