Server without obvious backup solution?

[mobycl1ck]

Hello STH
I was assigned to manage and administer a Linux server, being it a critical part of the infrastructure.The server was initially managed by an external firm, and now the server is on prem and it came that way.
Now, after initial assessment, I came to realize that the server does not have a "roll back" feature, only it has a backup script that backs up /home for all its users and /etc, using rsync to another physical server set up with SSH key.

But the server, as in the OS and all, dos not have a back up mechanism, although is fairly well scrutinized and hardened.
A few questions:
Can I install Timeshift as a intermediate step, just to make sure I can roll back the server, if something goes wrong?
What backup solution would be appropriate for this situation?
Do I have to check all cron jobs to see if there is still something going on, as a backup solution?

 

[MrGuvernment]

First question for me is, does this "need" to be a physical server? Could it be migrated to existing virtual infra, if you have any?

 

[mobycl1ck]

First question for me is, does this "need" to be a physical server? Could it be migrated to existing virtual infra, if you have any?

Fair question and fair point.
The server was bought and set up by the previous contractor and now it is on premises. It is for the time being a physical machine and maybe in time it will be virtualized.
But until then, that machine needs to be managed and used as it is.

 

[MrGuvernment]

Completely understandable. Get it in a good place now, look to make better when possible. Sometimes it can be a good reason to push for a migration if management has to approve, is ease of backups and management, vs relying on other methods.

Def as you noted, document all cron jobs, when they run, confirm logs that they are running and no errors. Timemachine can work, but be sure not to fill up any drives with them either..

 

[mobycl1ck]

UPDATE
Tried to replicate de server in virtual environment, and IRL. Installed Ubuntu , during installation i created a software RAID with two disks.
Everything went as it should. Installed Synology ABB and backed up the server. Booted the server with another live Linux and deleted the table partition on both disks.
Booted again with Synology ABB recovery ISO and restored from the image backup.

Seemed that everything went well, but ABB recovered the table partition and the partitons on a single disk, without the RAID settings.
Now, this begs the question: Do i need to create the RAID level again using mdadm before preceding with the actual recovery?
Well, i did it and Synilogy ABB wasnt able to restore correctly.What I am doing wrong here?