Hostgroup Summary In Outbound Check

[toodaly]

I would like to be able to send the hostgroup summary of various remote Nagios XI servers to a central Nagios XI server. I've been sending host/service information, but it's starting to get overwhelming. Is there another way to get summary information to a central server instead of using inbound/outbound checks?

Thanks.

[jolson]

It sounds like Nagios Fusion might be a good fit for you. Any interest? https://www.nagios.com/products/nagios-fusion/

[toodaly]

I think when I looked at Fusion before, it was an all or nothing approach. I need to be able to have User A see, for example, East Coast sites, User B see West Coast sites, and User C to be able to see everything. I've implemented this using contacts associated with certain host groups (sites), so when they log in, they only see the host groups they are contacts for. The Nagios XI documentation calls this Multi-Tenacncy. Can this be implemented in Fusion?

[jdalrymple]

So unfortunately there is no easy solution I know of for you. There is some internal unspoken intent to revamp Fusion so that it can better handle yours and the rest of the world's needs, but that's a ways off.

Is the problem the fact that you have to dual configure everything? While I've never done it myself I've definitely heard of instances where people have scripted pushing the appropriate configuration components back to the central server to eliminate the need for the double-input.

[toodaly]

No the main issue is using inbound/outbound checks with Nagios XI displaying passive check results from ~2000 hosts, ~7000 services from 40 remote sites. The administrator has to scroll through pages of service results for the site he/she needs or perform a search. I'd like to display the hostgroup summary at the central Nagios XI server (10 hostgroups per site, ~400 passive checks coming in vs 10,000 passive host and service checks. The administrator can then remotely login to the remote Nagios XI server to further investigate of one of the hostgroup summaries indicates a problem.

Another option I was thinking of is to create a service per host group and plugin that reports the most severe status of the hostgroup. The remote Nagios XI server would be the host that these services are associated with and just passively (inclusion only) send that host/service data to the central Nagios XI server.

[jdalrymple]

Nagios XI displaying passive check results from ~2000 hosts, ~7000 services from 40 remote sites. The administrator has to scroll through pages of service results for the site he/she needs or perform a search. I'd like to display the hostgroup summary at the central Nagios XI server

Sorry - it's not clicking in my head why you can't do that today. Can you elaborate, or am I just unaware of a shortcoming of the software. To the best of my knowledge there is no reason passive hosts/services can't be sorted any differently than active ones. As a matter of fact - I do it.

[toodaly]

Instead of passively sending individual service status results for individual hosts using inbound/outbound checks, is there any way to passively send hostgroup or service group status to reduce the amount of information on the wire? The administrator can then log into the remote Nagios server to further investigate any issues.

While I've never done it myself I've definitely heard of instances where people have scripted pushing the appropriate configuration components back to the central server to eliminate the need for the double-input.

Do you have any forum discussions or links concerning how to do this scripting to send status back to a central server?

[jdalrymple]

I'm understanding what you're asking now. The answer would be no I'm fairly certain since really the constructs of hostgroup status and such don't really exist. For example you couldn't get the status of a hostgroup (or even the name for that matter) from status.dat - it only really exists after the CGIs make it.

Do you have any forum discussions or links concerning how to do this scripting to send status back to a central server?

https://exchange.nagios.org/directory/A ... ly/details
https://exchange.nagios.org/directory/U ... 29/details

Additionally, while it's difficult to find anyone who's elaborated on their configuration I will point out that to create a passive check on a central server requires no more than the following:

Code: Select all

define service {
        use                              distributed-passive
        host                             somehost
        service_description              someservice
        }

So parsing your /var/log/messages for unconfigured things and creating them is fairly trivial. Then copying in hostgroups and such from other hosts is generally also pretty easy if stuff is organized well. Admittedly I've never fully automated an environment, but putting the pieces together seems like a point to point process for the most part.