Nagios Support Forum

This isn't really a support question (yet), but wanted to throw this out to see if anyone else has had issues with upgrading a system similar to what we're running:

Current configuration:
Nagios XI 2014R2.6 - 158 users (no LDAP/AD integration), 1600 hosts, 10,600 services with an offloaded MySQL database.

Desired configuration:
Nagios XI 5.2.3 with 600 users (with LDAP/AD integration) and the MySQL database on the Nagios box

Has anyone upgraded Nagios XI with a remote database? If so, how did it go?

Has anyone migrated an offloaded MySQL database back to a Nagios standalone instance?

We initially offloaded the MySQL database because of the potential load, but think we could safely move the database back to the main box.

The upgrade shouldn't care where the DB is located because it pulls the settings from the configuration files, I would make sure to perform an upgrade with this configuration in a test environment before upgrading your production server.

With that number of hosts, services, and users, my recommendation would be to NOT move the mysql database back to the XI server.

That being said, if you decide to go that route and do the upgrade I would strongly recommend that you create a ticket so that we can work through some of the things that may complicate this:

Since you are upgrading a pre XI 5 server it is still using a postgresql DB, a fresh install of XI 5 will NOT use a postgresql DB so if you intend on creating a new server you must install a pre XI 5 version of XI and then upgrade to the XI 5 version so that the postgresql DB will be utilized.

Migrating an offloaded DB back to the local machine is simple enough, essentially you:

Perform a SQL dump and import it into the local mysql database
Point the XI config settings for the DB back to localhost

Note: Previous backups of XI will still contain the settings for the offloaded server so if you try to restore it may fail, I haven't tested this.

How easy will it be to migrate all the checks from the old box to the new box (I really, really, really don't want a PostgreSQL database on the new box)?

We're also looking to migrate off of the NRPE agent (as much as humanly possible) to SNMP. Any tips on doing that?

Migrating the checks should be fairly easy (depending on your setup), however, you would need to transfer any custom plugins and re-install any custom packages that were required for the custom plugins.
- You could essentially follow this same process: https://assets.nagios.com/downloads/nag ... ios-XI.pdf with the config files from your current XI server.
- There may be some hiccups but it should be fairly straight forward, the problems usually arise when they aren't imported in the proper order (the import tool tells you what order at the bottom of the web page but people sometimes mix them up).
- You would also need to transfer all your /etc/mrtg/conf.d/*.cfg files (these are for bandwidth checks) and all your perfdata from /usr/local/nagios/share/perfdata if you wanted historical performance data.

For the NRPE portion of it, are you asking to migrate all NRPE based checks to SNMP? It really depends on what, and more importantly, how you are running those checks (custom plugins, scripts on the remote machine, etc)
- Again, this really depends on your environment, are most them setup through the wizards or manually?

Here is a little history as to what we're up against:

When we initially stood up Nagios, we wanted to use SNMP for all our checks (with custom scripts where needed), but got a LOT of push back - management and the sysadmin folks wanted to install NRPE, as they weren't comfortable with SNMP. So, we went the NRPE route.

We now want to install a redundant Nagios box (so we can fail over and do maintenance without service interruptions), but doing that would require all the monitored hosts have their nrpe.conf file modified to add the second Nagios server. Time has passed and folks are much more open to doing SNMP. Since we're going to have to touch each monitored host anyway, the decision was made as to "...why not just make the switch to SNMP..." wherever we can.

In addition, we have an older (Nagios XI 2014R2.6) and want the new features of XI 5.

We've got Nagios currently running on an HP ProLiant DL380p Gen8 (2 X Eight-Core Hyper-Threaded Intel(R) Xeon(R) E5-2660 0 @ 2.20GHz) & 192GB RAM - so I'm wondering if this box is beefy enough to run both Nagios and the MySQL database on the same box?

The big thing that I would be concerned about is you'll need to install net-snmp and set it up on the linux hosts. https://assets.nagios.com/downloads/nag ... g_SNMP.pdf

That machine is very beefy - I would assume that it can run both...that's a lot of resources.

Do you have any further questions regarding this?