Nagios Support Forum

Hello,

I have been setting up a NagiosXI server using the supplied VMware image. This was a 2011R1.1 image which has been upgraded to 2011R1.2, and then today I applied the 2011R1.3 update.
As I have been adding in switches and computers using the configuration wizard, the Nagios machine's CPU usage has been steadily climbing.
Currently, with:
# Active Host / Service Checks: 52 / 3219
The machine's CPU use is sitting around:
load average: 11.96, 12.10, 11.54
According to top.
Our license is for 100 hosts, but it seems like with 52 hosts I am maxing out capabilities of the supplied image.
Having this many hosts/services has also seemed to have the side affect of causing new configuration changes to take several minutes to apply. It makes applying changes quite a tedious process.

There is nothing special regarding this installation that I know of, aside from having Nagios Fusion installed along side XI in the image. We're eventually going to roll a 2nd server out at our 2nd location and us Fusion as a dashboard, but as of right now Fusion just as the local XI server located on the same machine in it. I was having the high CPU use before installing Fusion though, so it's not related, just thought I would mention I have Fusion installed too.

Check to make sure the database is in working order, take a look at the following doc:
http://library.nagios.com/library/produ ... i-database

Switch monitoring is a little more CPU intensive than some of the checks, but I don't think it should be running that high for only 3700 checks. How far apart is your average check interval?

How many services do you have being monitored? (It's a little hard to guess just given the number of hosts.)

One thing to keep in mind is that the VM is built with relatively low specifications to get started, but you are free to increase them. For instance, if you allocate four processor cores and 2GB of RAM to it instead of 1 core and 768MB RAM, the performance will improve considerably. (Assuming your host machine has those things to offer, of course.)

tonyyarusso wrote:How many services do you have being monitored? (It's a little hard to guess just given the number of hosts.)

One thing to keep in mind is that the VM is built with relatively low specifications to get started, but you are free to increase them. For instance, if you allocate four processor cores and 2GB of RAM to it instead of 1 core and 768MB RAM, the performance will improve considerably. (Assuming your host machine has those things to offer, of course.)

Ah, thank you.
I talked to my vmware guy and I shutdown the Nagios server so we could increase its resources. I went for 4 processors and 2gb of ram and that has alleviated the problem.
load average: 0.40, 0.62, 0.65
Also, the XI image only had 512mb as was. Maybe our vmware guy changed it when he added it in or something?
Not sure... either way, giving it more resources has fixed the issue.

Excellent, glad to hear it.