Windows 11 Pro, Installing Borg via WSL and/or Docker, and Backing Up Mapped Drives from a NAS on LAN?

[EnigmaSolved]

I have a Synology NAS which houses all of the data I want to back up (currently 17TB and growing, though initially will be starting with <2TB as that is what I am starting with at Rsync.net). My Windows 11 Pro machine has a fast SSD and i9 CPU, with 64GB of RAM, so I'm thinking that is more suitable to running Borg than trying to run Borg on my Synology (unless I'm going to upgrade my RAM on the NAS). Given that I'm a Linux newbie, I figure also installing Vorta to have a GUI will be helpful.

Any advice and considerations for installing Borg purely via WSL (Windows Subsystem for Linux) or via Docker (which is of course still on top of WSL), as well as recommended methods for mapping my network shares in whichever method?

~16TB of my current data is video files of varying kinds and sizes (ranging from 500MB to 50GB). That data does not change, though more will be added (eg, from ongoing videography recordings). I am very new to Borg, but I'm wondering if I'm thinking I might eventually backup all of this, is there any benefit to separating out my data somewhat? For example, having all of my videos in the same repository(?) so there's deduplication among those files (though there should be very few duplicates), but then having substantially different data (eg, text files, code files, Excel spreadsheets, etc) in a separate repository(?) which are only deduped among themselves?

I'm just thinking that some data I'll want to be backing up frequently (maybe multiple times/day), whereas other data (eg, videos) can be once/day. Though the videos will be so much larger, so those backups will run much longer. I'm hoping there's ways to prioritize backup sets such that the more frequent backup set can interrupt (or run concurrently to) a backup set of larger files that will take much longer. I haven't gotten into the weeds of the docs yet to figure out if/how that might work.

Thanks in advance for any advice/guidance on any of the above! :)

[ThomasWaldmann]

Any advice and considerations for installing Borg purely via WSL (Windows Subsystem for Linux) or via Docker (which is of course still on top of WSL)

I use neither of these, but WSL is said to be similar to Ubuntu, so you could follow the borg documentation for Ubuntu Linux.

~16TB of my current data is video files

Make 1 or more separate repos for these. Running "borg check" is slow for huge repos.

but then having substantially different data (eg, text files, code files, Excel spreadsheets, etc) in a separate repository(?) which are only deduped among themselves?

Yes, have a separate repo for these. It's also about retention time (borg prune), faster borg check, more frequent backups, etc.

I'm hoping there's ways to prioritize backup sets such that the more frequent backup set can interrupt (or run concurrently to) a backup set of larger files that will take much longer.

With borg 1.x, you can only run 1 operation per repo at the same time.

[EnigmaSolved]

Thank you so much for your responses -- very helpful!

Given that borg 1.x can only run 1 operation per repo at a time, and that my dataset makes sense to have multiple repos (eg, one repo for videos, one for smaller files, etc), it sounds like I could have 2 different borg operations running concurrently, so long as they are going to separate repos.

However, since most of my repos (probably all) will be hosted at Rsync.net I am wondering what the recommended procedure is for how to pause a longer backup (eg, videos) that may or may not be currently running, in order for a higher priority, more frequent backup to run, and then resuming the paused backup (if it was previously running) once the quicker, more frequent backup is complete?

Also, I should mention I am now leaning toward upgrading the memory in my Synology DS1522+ to 32GB and running the Borg with Borgmatic package on that (instead of going the WSL route). I'm hoping the AMD Ryzen R1600 dual-core (4-thread) 2.6 GHz (with max boost up to 3.1 GHz) will be sufficient, in conjunction with more ram (my assumption is ram is the more important thing). Regardless, I'm not totally clear yet from what I've read thus far as to how to accomplish the above (pausing/resuming a lower priority backup in relation to a higher priority backup) in Borg and/or Borgmatic. Please forgive me if it is super-obvious somewhere! 😆

[ThomasWaldmann]

There is no pausing.

You need to terminate a running born (backup) process if you need to do something else in the same repo.

If you later restart the backup, it will re-use the already transmitted / checkpointed data.

[EnigmaSolved]

Okay, thanks much -- glad to know I wasn't missing an obvious command somewhere!