nagios replication

[nshahorchard]

We're looking for a solution (preferably within nagios) that would allow us to have multiple nagios servers being replicated and probing the master. In other words, one stop shop for configurations and depending on the location, it will copy the configs to that location. Wonder if nagios fusion would be something that would do this? Any other suggestions would be great.

What we're looking for:

Nagios replicated to a remote site
Nagios config entered in one central place and then copied over

[abrist]

Nagios Fusion is a central view for multiple Nagios instances (XI and core).
Do you actually want to replicate the entirety of the nagios config to a remote site, or just the checks relevant to the remote site?

[nshahorchard]

Ahh I see, I thought nagios fusion was a centralized place to create new checks within multiple nagios instances.

As far as your question, I was hoping for both options. If you can point out what I could use for replication and what I can use to deliver just the relevant checks to the remote site that would be very helpful. I don't know of anything that would accomplish this.

[slansing]

What you could do here is install Nagios Core on the remote site, "or use puppet to push it over and copy your configuration files over." Or, an easier way would be to push the check results from your nagios servers to a central one via NSCA. The way this is accomplished is by forwarding state changes and status details to the central servers passively. To find details on setting this up please visit this document and search for "Central Server vs. Distributed Servers".

http://nagios.sourceforge.net/docs/nagios-3.pdf

[nshahorchard]

So I read the following regarding NCSA; please excuse my ignorance since I've never implemented this.

http://nagios.sourceforge.net/download/ ... _Setup.pdf

Now as far as the setup, would this be configured on the remote server? Once this is setup, would this replicate to the primary server? How does this work behind the scenes?

[slansing]

Once you have your remote servers set up and receiving check results from their specific checks, they can forward those results up to the main Nagios server which will include all of the hosts and services from each remote Nagios installation.

And, you do not need to excuse yourself for ignorance as I believe learning does not count as ignorance, rather.. ignoring the learning process would be..

[nshahorchard]

1. Understood that part; however, in essence the host needs to configured initially on both remote and primary server. NCSA basically just fowards the results from the remote to the primary so that the primary server doesn't have to do all the heavy lifting. Is that correct?

I think this still defeats the purpose of having to configure twice initially? Unless I am mistaken.

2. Ah, reading up a bit more it seems like the NCSA works on client machines. Whereas, you have a server and a host, the host does the checks and reports it back to the server. I don't think this would work in an environment where I have (2) nagios servers?

3. Have you ever worked with nagios using chef? Or only puppet?

[sreinhardt]

Either chef or puppet will work fine if you wish to run duplicated nagios servers. As for your idea on nsca, yes normally it is considered something run on a client machine, however it absolutely can be run on a standalone server and used to check other devices in that network, then forward up the results to another nsca server, or directly to nagios.

Also in the future if you could, please edit your past post to include future comments if no one else has replied. Multiple postings actually push you lower on our "to be responded to page" as you show as a newer post.

[nshahorchard]

Just FYI, I decided to install 2 instances of nagios (internal and external utilizing chef). I also implemented a tool called thruk for that 1 interface feel and also allows you to make minor changes.

[abrist]

great! should we close the thread?