Nagios Support Forum

Hi,

I've been asked to find out out what these messages mean in the event log and if they are anything to be concerned about?

This is a warning designed to warn if you're doing a huge number of checks that may overload the system. Here's how it calculates this number.

Linux PIDs have a maximum of 65535 processes (threads) running at once. Processes may fail to start if the processes limit is reached.

I must ask: How many hosts and services are you checking? How often are those checks going out? Every minute? How's the CPU and network usage on the Nagios XI machine (run top)?

You could greatly lessen the load on the system by increasing the delay between checks. It might even be a good idea for checks like disk usage and SSL cert expiration to have a much longer delay between checks -- say 1 hour to 1 day.

We have 1508 hosts and 13566 service checks.The majority of the checks are every 5 minutes. Some checks are once a day.

We have an offloaded database and the checks are offloaded to 3 Mod-Gearman worker servers.

Since you're using Mod Gearman, the warning isn't about a problem that can happen on your system; Mod Gearman splits the load between multiple servers, but the warning only applies if all checks went out from the same server.

Warnings are just warnings after all. Your system configuration is fine - it doesn't seem like it'll come anywhere near the process limit.

That's good news.

Is there a way of increasing the limits, like the warning message says?

My line manager would like to know if there is some way to make the message disappear from our system.

The calculation is invalid because nagios doesn't run all the checks at the same time and that message should be ignored (this was told to me by the dev that wrote that RLIMIT_NPROC check into Core when I asked him about it).

There isn't currently a way to remove it from your system without increasing your nproc limits to be higher that the value listed, for example, you can add these to your /etc/security/limits.conf:
Then reboot the system to pick up the limit changes and see if that warning is gone.