Nagios Support Forum

I know Nagios XI is setup to have a primary server and with licenses you can have other servers monitor parts of your network and report back to the main server. My boss is wondering if there is an issues with having two main servers, up and running and performing checks all the time. He does not like the parent child setup although I tried to sell him on that. He definitely didn't care for using a Nagios instance to monitor if the primary server goes down, so I have to work on the extra server route.

Thanks

Having two separate servers is not a problem, it just means you're doing double the monitoring, double the amount of traffic to the devices being monitored etc, double the notifications.

Keep in mind this will require the purchase of an additional XI license.

Yep we bought the licenses, then my bosses got changed and the new boss was looking at it a different way. My thought was more inline with how it was designed. If you put a separate server in a location, then it monitors that area and reports back to the main server. Seems like a good deal for load balancing too, but I'm losing that argument

I saw an article that says for larger monitoring environments >1,000 hosts and services, performance tuning may be necessary. We're at 3,372 monitored hosts and services. At what point would you think the need for additional servers, monitoring their own section of the environment and reporting back to the main server, would be necessary?

That may have been in some of the documentation as best practices, but I don't recall seeing it.

Thanks

EnvBroker1 wrote:I saw an article that says for larger monitoring environments >1,000 hosts and services, performance tuning may be necessary. We're at 3,372 monitored hosts and services. At what point would you think the need for additional servers, monitoring their own section of the environment and reporting back to the main server, would be necessary?

Generally Nagios starts to have problems when the total host and service objects is > 20,000.

Have a look at this presentation on Nagios XI Best Practices:

https://www.youtube.com/watch?v=6WlZrG-_sAI

The presentation includes some performance tuning like enabled a RAM Disk.

What is your primary goal for have 2 XI servers?

I suggest you look at Mod Gearman if you just want to execute the checks at remote locations. It can offload the checks while still only having one XI server.

EnvBroker1 wrote:He definitely didn't care for using a Nagios instance to monitor if the primary server goes down, so I have to work on the extra server route.

Having a second XI server monitor the primary one is a good practice. If XI dies, it's not going to tell you about that, or any other critical things like the server room temperature has just exceeded a threshold. I cover this at the beginning of my presentation. You can use your DR license for this purpose so it doesn't cost you anything.

Does this help?

Thanks for the info Troy, sorry I haven't responded sooner.

The two (full) server setup was what my boss was looking at. I explained the whole monitoring the Nagios instance, but he wanted some more options. At this point and time I think he's leaning towards us just monitoring the primary server from another server like you guys always explain.

Thanks again and you can close this thread.

Eric