Sending Data Between Nagios Servers

[optionstechnology]

I have a number of internal servers which monitor our network devices and client servers that manage the clients servers

Problem is the client servers cannot ever directly connect to the network devices - but the two nagios servers CAN see eachother

So I need to pass check results + perf data from one Nagios Server to another

I checked out NRDP, it seems to do the trick - I'm just wondering how you guys normally use it to get realtime data?

Do you set it up as like an event handler? Or is it just a cron job that runs every 5 mins?

[scottwilkerson]

If you use the NRDP using the inbound/outbound transfer in the Admin menu I would say that would be the recommended method.
https://assets.nagios.com/downloads/nag ... ith_XI.pdf

[optionstechnology]

So does outbound check transfer just send everything across and only stuff with a matching hostname shows up?

[scottwilkerson]

It sends everything that isn't filtered by the "Host Name Filters:" regex filter.

For example this means doesn't begin with localhost

Code: Select all

/^localhost/

so it would send everything except hosts that start with localhost


Then, on the receiving server, if the host /service isn't configured yet, it will show in Admin -> Unconfigured Objects

[optionstechnology]

Awesome, ok I've got that setup and running now

Two questions-

1. How do I send it to more than 3 servers? I have 30 Client Nagios servers that I have to send data to

2. How do I specify which server to send which data to? Or do I have to send it all to all of them?

[scottwilkerson]

I'll answer #2 first, they all have to go to each.

For #1, I've never had this question before because usually people are doing exactly the opposite, they are performing the checks on numerous XI server and want to send up to a single central server (sometime 2-3 if they are also sending to their test and development environment) for a unified overview.

The short answer is we could consider making it more as a feature request, but if you are sending all the data to all the servers, begs to ask the questions, why not just let the users see the data on the main server that is sending the data?

[optionstechnology]

We would have to grant the users access to the main server from their network to ours - which would breach our security

Plus the only thing stopping one client from seeing anothers data would be the username control on the nagios box

[scottwilkerson]

We would have to grant the users access to the main server from their network to ours - which would breach our security

Plus the only thing stopping one client from seeing anothers data would be the username control on the nagios box

Is each clients Nagios XI performing their local monitoring? If their monitoring was being performed on their local XI server they could just log into it and see what they need, or is their something on the main XI server which they need access to see?

[optionstechnology]

Yeah pretty much - we monitor the local client servers on their local client Nagios server

And we monitor network devices on a central server which the customers cannot have access to

I want to push data outward from this central server to the outlying ones so each customer can see some of the stats of the network devices - this way the clients have no inward route into the main network


I'm happy to script something to solve it, just not sure what the best way to go about it is

[cdienger]

Event handlers calling one of the send_nrdp scripts found in /usr/local/nrdp would give users some insight to network devices. Having a designated handler for each site would also allow you some control over what data is sent where. The draw back to handlers though is that they're only triggered by problem states and recovery.

https://assets.nagios.com/downloads/nag ... ios-XI.pdf