Backing Up Lightroom Classic Photos to the Cloud

I have 35,000 photos and follow a similar scheme with a caveat. I live in Canada means expensive Internet and I don't have an unlimited plan so backing up all my photos can be costly. I've come up with a 2 step process.

For the offsite:

• I refresh off-site physical backups to my daughter's and our cottage quartlery.
• I come from a computer background and wrote a program to find all my 'keeper' photos (rating >= 1 star) and copy these to the cloud. If anyone knows Python, and is interested in the program, let me know.

For the cloud, I use Amazon Web Service (AWS), which is cheap, but that requires me to use the Cloudberry Desktop backup program. As an FYI, if you choose to use Cloudberry, don't use any of it's special features like compression. This makes you bound to Cloudberry because other 3rd party programs that work with AWS cannot read the Cloudberry created containers.

I would just comment: whenever backup happens there are two competing aims in play.

The first aim is to maintain a "mirror" of your current primary data, as up to date and identical as possible, that matches and reflects all of your ongoing intentional additions, alterations and deletions.

The second aim takes into account that unintended changes can happen too. So files - many or few - may have been deleted, or corrupted, or changed in some unwanted way whether accidental or malicious (ransomware, say). 

And you didn't realise quickly enough, to block those updates from getting automatically mirrored to all your backups too; the very fact they match the primary data, would be what makes them useless to you in this circumstance.

The answer is either to maintain incremental backups that permit going back in time; or, to deliberately disconnect or prevent some backups from getting updated at all, except when you want that to happen.

My suggestion is: use cloud backup or local backup to very regularly update a simple mirror of all your "presumed good" data, to protect you as much as possible from losing work in the case of a sudden local hardware failure. And separately from that, at wider intervals, maintain a circulating series of physical backups - that are then disconnected and stored securely away from your main computer. These would serve as a redundant backstop: deliberately out-of-date but "presumed good" data snapshots, to cover the scenario where both your current data and its cloud based "mirror" are found to be "bad": to embody a lot of unwanted change. It may easily have taken a week, or a month, before this became apparent. If your backup software allows an incremental method, one external drive might permit the retrieval of old data states from say 1, 5 and 9 weeks ago - the next drive from 2, 6 and 10 weeks ago - and so on.

There are plenty of cloud backup services. Each comes with different instructions. I use Carbonite as cloud backup, which actually detects new or changed files, and then performs the backup automatically without human interaction.

You would be wise to add to your 3-2-1 system (which I commend you for following) to make your backups automated, so they happen without human interaction. In my opinion, no backup system is operating at its best with human initiated backups (why? because humans forget, humans get busy, humans get lazy, and backups don't happen regularly).