Hello,
I am new to Nagios and I have a little problem.
Our company has 2 facilities, that aren't connected in any way, only over the internet.
My question is: Can I host a Nagios-Server at my facility and host a smaller one in the other facility and the smaller one reports everything over the internet to the main server?
Or are there other ways that I could achieve my goal of having only 1 real monitor for both facilities?
I hope you can help me.
Thanks
Ben
Monitoring 2 facilities with the same server
-
Ben Morgan
- Posts: 1
- Joined: Tue Jul 10, 2012 2:33 am
Re: Monitoring 2 facilities with the same server
There are several ways you could go about this and the right way is going to depend largely on your companies security policies and DMZ design. I'll outline a couple of options:
1. Configure two nagios installations (1 at each site) and use the Nagios MNTOS plugin which will give you an overview of both. This has the downside of needing to maintain two Nagios servers.
2. Configure two nagios installations (1 at each site), create passive services on your HQ Nagios and have event handlers on the remote Nagios server that relays the information back to the HQ. This has the downside of potentially having significant configuration overhead but does centralize you to a single authoritative nagios source.
3. Configure one Nagios install in the HQ with a NRPE proxy at the remote site, basically all checks for the remote site will hit the NRPE proxy and from there it will attempt to check the required devices. This requires that you can traverse the DMZ in some fashion and may also limit some of the checks that you are able to perform (particularly ones where you need to get a bit... creative).
4. Configure one nagios install in the HQ and use NSCA/SNMP to report information back to the HQ, this will require that all of your services are passive. The idea of potentially updating an entire site worth of passive checks makes me itchy.
There are a few other ways you could go about it, but this should give you some food for thought. If I was setting this up, my preferred option would probably be to use a VPN to an NRPE proxy where I would have a duplicate of the libexec folder... I would then use macros and arguments as necessary in the configs to dynamically form the checks to run. For passive stuff such as SNMP traps I would have to install the translator and NSCA to send that info back.
1. Configure two nagios installations (1 at each site) and use the Nagios MNTOS plugin which will give you an overview of both. This has the downside of needing to maintain two Nagios servers.
2. Configure two nagios installations (1 at each site), create passive services on your HQ Nagios and have event handlers on the remote Nagios server that relays the information back to the HQ. This has the downside of potentially having significant configuration overhead but does centralize you to a single authoritative nagios source.
3. Configure one Nagios install in the HQ with a NRPE proxy at the remote site, basically all checks for the remote site will hit the NRPE proxy and from there it will attempt to check the required devices. This requires that you can traverse the DMZ in some fashion and may also limit some of the checks that you are able to perform (particularly ones where you need to get a bit... creative).
4. Configure one nagios install in the HQ and use NSCA/SNMP to report information back to the HQ, this will require that all of your services are passive. The idea of potentially updating an entire site worth of passive checks makes me itchy.
There are a few other ways you could go about it, but this should give you some food for thought. If I was setting this up, my preferred option would probably be to use a VPN to an NRPE proxy where I would have a duplicate of the libexec folder... I would then use macros and arguments as necessary in the configs to dynamically form the checks to run. For passive stuff such as SNMP traps I would have to install the translator and NSCA to send that info back.