I have read about the Archive node and it is slightly different from the usual backup scenario. I am trying to find out if we can do a classic backup that moves everything out of the "grid" but allows us to recover it later altogether.
This is to cater for a Disaster scenario when, for example, an engineer has deleted all buckets or even more, destroyed a full grid storage at all. (believe me, it happens 🙂 )
The Tivoli Manager is from IBM, I don't know, are there any better "modern" 😉 solutions to solve the problem?
Perhaps, an equivalent of something like an SQL DB backup - a single .bak file that allows you to recover the whole Database. I wonder if something similar may be in the product portfolio. Or, for example, we use DELL Avamar Backup, which supposedly works with AWS S3 backup, but I am not sure if it supports the StorageGRID S3 implementation.
Something that could wrap all objects altogether and write them to another storage. And if needed recover them all.
Object Storage doesn't lend itself to the concept of backup.
It's not meant to be back-up'ed, but if you need a copy in different pools, you setup ILM with 2-copy policy and ensure one copy goes to each pool.
If you need a copy outside of StorageGRID, copy or download data to the other place (and that other place could be another StorageGRID). You can use NetApp CloudSync to make on-demand copies, or built-in CloudMirror  to queue changes and replicate elsewhere, or other tools to do the same, or even make a copy on the app layer so that you don't have to make copies later. It's really meant to be used in such a DIY-way rather than one "storage guy" being responsible for everyone's data like with enterprise NAS or SAN storage.
you can protect your sotragegrid/object data in a few different levels:
If you want to protect the data from physical threats then you have ilm rules to replicate data to different nodes and even different sites so continuity can be achieved with data replicated to different site/sites and retrieved from them in case of a disaster.
If you’re worried about logical damage/deletions then clients can activate from their side the versioning options which will keep objects even after they’re deleted for recovery.