Nagios Redundancy option

[nms]

Hi,

We are planning to set up a geo-redundancy setup for Nagios XI.

At present we have a Nagios XI in location1 which is a virtual machine (CentOS) monitoring ALL host/services for all locations.

Ideally, we are planning for another instance having the same data and probably run in synch, thus if one Nagios instance is dead, the other takes over.

Option:

Nagios XI in loc1 monitoring everything, Nagios XI in loc2 synched with Nagios XI @ loc1
Ideally, I was thinking in having some sort of VIP IP address that both Nagios server points to (sort of cluster), so it's transparent to the user (i.e user logs in always to the active node).
The license should contain 3 licenses which you can use as a standby?
How'd you do that? Can you just apply the license on the other Nagios XI instance?
How to best synch the 2 Nagios instances?

Rgds,

Matthew

[npolovenko]

Hello, @nms. Yes, you can use the same license to activate the backup XI server. As long as both servers are not actively monitoring services and hosts at the same time you should be fine.
Would your backup XI server in the second geolocation have access to all devices monitored by the XI server #1? Would you use a VPN?
Nagios XI doesn't have built-in supported functionality for redundancy. But I've seen users implement such solutions in their environments using custom scripts to a)backup production XI server and restore it on the backup XI server, on a daily/weekly basis b) to activate nagios monitoring on the backup server #2 if the first server goes down.
We also have a KB article with general insights and recommendations:

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

But the easiest way to make sure the production XI server is running is by setting up a backup XI in the second geolocation and monitor XI server #1 from the backup XI server. We have XI monitoring wizard for that. That way you will be notified if something wonky going on with the production server. Also, set up daily/weekly automatic backups on the production XI, or on a VM itself.

[nms]

Hi,

Quoting - "Yes, you can use the same license to activate the backup XI server. As long as both servers are not actively monitoring services and hosts at the same time you should be fine."

How do you apply the license and "tell" Nagios to be the standby node?

As ideally there will be restoration on the 2ry server on a daily basis.

Also, there is the other option, but I kindly require some feedback:

Active - Active, one with Nagios XI (licensed), the other with Nagios Core.

Can I somehow use the XI backup to load up all current host/services on to Nagios Core?

Rgds,

Matthew

[ssax]

How do you apply the license and "tell" Nagios to be the standby node?

Disable active checks and passive checks on the standby server so that it's not actively monitoring, or just shut down the nagios service.

See here for the license entitlements with better clarification, if you have any further licensing specific questions please reach out to [email protected] as we do not have access to that information:

https://support.nagios.com/kb/article.php?id=145

I would generally just do an XI backup -> Transfer -> Restore and then re-IP the new server to the old server OR follow the After the Restore section of this guide:

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

You could also setup SSH backups from the XI server to be stored on the standby server by using Admin > Scheduled Backups and then they would be immediately available for restore.

Active - Active, one with Nagios XI (licensed), the other with Nagios Core.

There is no automated process to do this but since Core only uses the config files and not a DB by default, you would need to copy over the config files from /usr/local/nagios/etc that are a part of Core and restart the nagios service on the Core machine.

Historical data and all that info would not be transferred, that would be a whole another process that would be simplified by just doing Active/Standby. If you want HA currently you'll need to write your own solution or use DRBD (links in the original guide npolovenko send).

You could also look at using Admin > Outbound Transfers but you would need to setup NRDP on the Core server to accomplish that.

[nms]

Thanks, you lock this thread.