Hello all,
We are a UK based MSP and have need to get remote monitoring running across most of our clients from small businesses of around 30 workstations and 2-5 server all the way to 1000+ asset networks in schools. We have done our research and from what we can see Nagios seems to be one of the most flexible and scalable solutions available.
I have however run up against a snag. We have a demo server in our office here running XI in a VM and its monitoring the LAN here with no issues at all, I have added a single exterior host to the set up (one of our PC's at another site) we did this using NSClient++ and a port forward to send the traffic to the local IP of the workstation, which works fine. However we are unable to add another host on the same LAN to the XI instance running in our office as the traffic is obviously being directed to the one workstation in the forwarding rules .
My question was originally going to be "can we run a nagios core instance inside each LAN and then aggregate the data on one central instance in our office" which we can with Fusion apparently. However i have been advised by the Nagios sales team that we may be able to achieve the same thing using one XI server in our HQ using passive checks, I have dabbled with them but i honestly have NO idea where to start with those.
For the two largest networks (1000+ assets) we would probably want to put a dedicated XI instance inside the LAN and is possible send the data to our core XI process at HQ (if that is even possible) and have all the other clients report directly to our core XI instance via passive checks.
So my questions are thus.
Can you have hosts added to XI behind a NATed connection using only the one XI instance at our HQ?
Can you have one XI instance on on LAN1 link up to an XI instance on LAN2 via ADSL and have the process on LAN2 (for arguments save our HQ) display all the data from the process on LAN1 ?
And finaly, can you use a Nagios Core on LAN1 to connect to an XI instance on LAN2 via ADSL and display the combined data?
This is obviously the research stage of our work , but the end plan is to set up a NOC for all assets on all our managed networks. I will admit openly, i am no linux sysadmin I have a working knowledge of centos but i am somewhat our of my depth on the Nagios Core front (all the config's give me nightmares)
Thank you in advance for your time.
NagiosXI monitoring multiple LAN's over NATed ADSL
Re: NagiosXI monitoring multiple LAN's over NATed ADSL
I've seen this question a few times and having never set it up my knowledge is a bit sketchy and hopefully one of the devs can answer better than I can but here's what I can offer in regards to answering your questions:
1. I believe you can by doing a shifty. You can use NRPE (or NSClient for windows) as a command proxy server. You will execute an NRPE check to the proxy server that will execute another NRPE check to the target internal LAN1 server... there's probably some draw backs but I don't see why it won't at least fundamentally work.
2. No idea on this one, there are some distributed Nagios plugins but I don't know if they are purely for load balancing or if they have capabilities beyond that I don't know about.
3. See above.
Hopefully that gives you something to at least experiment with
1. I believe you can by doing a shifty. You can use NRPE (or NSClient for windows) as a command proxy server. You will execute an NRPE check to the proxy server that will execute another NRPE check to the target internal LAN1 server... there's probably some draw backs but I don't see why it won't at least fundamentally work.
2. No idea on this one, there are some distributed Nagios plugins but I don't know if they are purely for load balancing or if they have capabilities beyond that I don't know about.
3. See above.
Hopefully that gives you something to at least experiment with