Hi Guys,

 

Log time user and first time poster, I have a questions to ask the users of Unraid.

 

** Disclaimer ** This is all theoretical and not something that is implemented or even close to implementation.

 

I'm looking at trying to achieve some kind of redundancy from manual or automatic file deletions as possible. So for example if x amount of files where deleted by accident on the server then I would have them replicated somewhere else to restore them some with ease.

 

For example if I had 6 x 4TB disks, of which I put two in Duel parity. Two of these disks I would have on User share 1 and two disks on user share 2.

When ever I add data onto the server it would be added to User share 1, then on a monthly basis a rsync job would run to replicate all changes from User share 1 to User share 2 therefore creating a "backup" copy. Ofcourse its not a backup but more of a mirror of data.

 

Other than the problems of multiple disks failing or the server failing does anyone see any issues with this kind of setup?

 

I know I'm possibly missing some obvious reason as to why not to do this but my mind is blank at the moment.

 

Thanks

AlexP

17 minutes ago, AlexP said:

I'm looking at trying to achieve some kind of redundancy from manual or automatic file deletions

Recycle Bin plugin, only helps with deletions over the network.

 

User Scripts plugin has scheduling.

 

In general though, you should have backups of any important and irreplaceable data on another system, possibly offsite.

 

25 minutes ago, AlexP said:

Other than the problems of multiple disks failing or the server failing does anyone see any issues with this kind of setup?

There certainly have been some disaster stories here on the forum.

5 minutes ago, trurl said:

Recycle Bin plugin, only helps with deletions over the network.

 

User Scripts plugin has scheduling.

 

In general though, you should have backups of any important and irreplaceable data on another system, possibly offsite.

 

There certainly have been some disaster stories here on the forum.

 

I'll check out the Recycle Bin plugin, I did not know something like that existed.

 

Generally the data I am storing on the server although important is not irreplaceable. Everything that is irreplaceable is stored in a 3-2-1 fashion. However at this point in time cost is more of an issue than the importance for backups so I'm looking for a medium term compromise on having a small amount of redundancy in the face of data loss vs cost. Doing it all on one set of hardware seems like a good compromise at this point in time. But maybe I'm looking for someone to talk me out of it haha.

 

Thanks

AlexP

The issue you face is, how can you be sure that the event which deleted your files to begin with did not affect the mirrored copy?  If it was simple user error you should be good (unless the sync has run and replicated the deletes to the mirror).  But if it was malware then you'd better have the mirror heavily secured.  And if it was a hardware issue, then who knows?

 

So your strategy would work to a degree, but it will be quite limited in what real world problems it actually protects you from.  Since it seems like your 3-2-1 strategy doesn't facilitate fast/easy restoration, maybe it's worth looking at that?  For instance, I have an 8TB drive in a portable USB3 enclosure that I keep locked at work.  Same basic principle, mirror the data with a Robocopy script, but capable of recovering from more scenarios than colocating the mirror within the same array.