nagios high availability

[benhank]

Hey fellow nagiosians!
i am asking for a little guidance\ advice or pointers regarding Nagios high availability solutions.

we have about 1543 hosts and 5079 service checks running.
I have one server running nagios, and two spares that can be tasted for high availability.

My question is what are the high availability setups are you guys using.

By high availability, I mean A setup where there is no interruption of service when a problem arises. And also a system in which nagios has minimal slowdowns do to high cpu utilization.

my machines are running dual 16 core processors 1.5 terabytes of hddspace and 8 gigs of ram.


Thanks for you impute and advice in advance.

[scottwilkerson]

We do have a document on High Availability in the Nagios Library.

[benhank]

i read it, but i wanted to know how folks have set up their own HA solutions.

[scottwilkerson]

This really is a high level overview of the types of things that are done for high availability situations.

To get much more detailed than this would need specifics of you setup and infastructure, which would bring us outside the scope of general support and into consulting.
http://www.nagios.com/services/consulting

[benhank]

thanks again scott. I never thought about it that way.

[paul.jobb]

We currently have 1500 windows hosts with approx 12,000 service checks,primarily check_wmi_plus, running at 5 minute intervals. We are using mod_gearman as the distributed broker instead of DNX and it seems to working fine. I have one nagiosXI vm appliance(2 vcpus, 4gb ram), 5 vm's running centos 5.7 as gearman workers and a dedicated VM as the mysql database server, system response is very good with this configuration.

[benhank]

hey thanks paul! this is what I was looking for , other options I would never have thought of. Thanks a lot sir!
If I may ask, did you guys use a different set up before you switched to what you are using now? and if so what caused you to change?

[jbennett]

We currently have 1500 windows hosts with approx 12,000 service checks,primarily check_wmi_plus, running at 5 minute intervals. We are using mod_gearman as the distributed broker instead of DNX and it seems to working fine. I have one nagiosXI vm appliance(2 vcpus, 4gb ram), 5 vm's running centos 5.7 as gearman workers and a dedicated VM as the mysql database server, system response is very good with this configuration.

Is there a reason you decided to go with mod_gearman over DNX?