Nagios Support Forum

Good afternoon, I'm looking for some insight to hopefully resolve an issue we're experiencing.

Whenever we perform a Log Server upgrade (i.e. 2.1.1 to 2.1.2), I like to take a snapshot of the nodes in vCenter in case we need to roll back the upgrade. The issues with this approach lately is that each snapshot in vCenter is consuming 9TBs of disk space and it's becoming less viable for our environment. We have 3 nodes in our cluster.

I'm curious, if we did have a problem with an upgrade, is there specific subset of data/directories would could focus on backing up and still be able to rebuild a node, restore the data from backup and still have a fully functional cluster?

Thank you.

If I recall correctly, you have Log Server snapshots running to a repository on another server. Is that right? If that's the case then those snapshots should cover you most of the way. You might take an immediate snapshot just prior to the upgrade, and that would cover the logs up to current.

The above will make sure you have a copy of the data in the Elasticsearch database. Configurations like accounts, credentials, dashboards, etc... would be included in a system backup. A backup should be created automatically everyday, but you can generate one manually as well from the command line:
They are stored in /store/backups/nagioslogserver/

You can find details on backing up and restoring a system in https://assets.nagios.com/downloads/nag ... Server.pdf.

So, to clarify, the backup and the snapshots that occur nightly are equivalent to the vCenter snapshot I run prior to upgrading the nodes to new versions of Log Server?

Everything you would need to restore a node or cluster would be found in the NLS system backups and nightly snapshots(minus the current day unless you force it with the command provided).

Great, thank you for the clarification. You can lock this up.