Hi, before I can fully commit to moving my 2TB of data (personal and business) from Mac OSX to Fedora and finally say (a long overdue) good bye to Apple, I need to know if I can secure my data as easily, or at least if I can do it in some fashion.
I have used TimeMachine on Macs for many years, I have two TM drives constantly connected and backing up (giving me two in case one drive fails). I also take a CarbonCopyClone every 3-6 months which is stored in a fire safe.
How would I go about doing something similar with Fedora, is it possible?
I suspect ongoing backups are possible, but I’d really like to know if there is an equivalent to a full clone (bootable) so if I lost my Fedora laptop with 20 years of data on, I could buy another and clone from my backup drive to make it identical to the one I lost (excluding files edited since last clone). Is there a way to do something like that?
There are many tools to do exactly what you describe. ‘rsync’ would easily do the onboard incremental backups to your externally connected drives, and clonezilla is one of several to do the full system backups for off site (fire safe) storage.
I am sure others may chime in with alternatives as well.
see also Recommended daily backups and You searched for backup - Fedora Magazine
the key is to be able to restore a backup, which, depending on the backup method, isn’t all that trivial…
I personally only backup my /home (different methods, different drives and different locations) and don’t care about the system itself
Thanks very much, some great suggestions there.
I stupidly forgot to say, YES - GUI ONLY is what I am looking for! It’s hard enough to move to Fedora without introducing CLI stuff any more than necessary
I will have a look at the above suggestions, and agreed, the system (OS) itself doens’t ‘need’ to be backed up. What I always loved about Carbon Copy Clone though, is I could actually boot from it from a spare machine and manually access a working OS with all my files in situ, has been very useful from time to time.
I think I may actually keep a spare old iMac for a while, just for that purpose, so if anythin goes wrong or I am struggling with Fedora, I can boot back to comfort and find any files I need.
To have something like that for Fedora would be awesome, but not ultimately essential.
Thanks again, I am so impressed with the great help given on these boards. I have a warm and fuzzy feeling about moving to Fedora and away from Apple (and their arrogance a55holes on their forums!)
As mentioned by Flo, Timeshift is very similar to macOS Time Machine. You won’t get the neat animated scrolling back in time experience, but it’s functionally the same thing.
Timeshift can work with Btrfs too, which is the default filesystem on Fedora. The last time I used it, Timeshift only supported Btrfs subvolumes that were prefixed with “@”, but the default subvolumes on Fedora are “root” and “home”. If you want to use Btrfs with Timeshift, you’d have to reinstall Fedora and use the Advanced Partitioning option to rename the default subvolumes. Rename “root” to “@” and “home” to “@home”.
Alternatively, there is a way to create new @ and @home subvolumes from snapshots of “root” and “home” without having to reinstall Fedora, but this involves use of the command line, experience with Btrfs, and the potential for data loss. As you mentioned, you’d prefer to keep command line usage to a minimum, so this might not be the best option for you.
Thanks Jeffrey. Whilst that probably sounds trivial to most Fedora users, to me that sounds like quite a headache and probably prohibitive (the “reinstall Fedora and use Advanced Partitioning” thing).
I take it there are not other options which can do similar without the need for that?
This is my preferred technique. Just copy the /home folder. The system itself gets a clean install occasionally.
If you sync folders with cloud storage just make sure it’s inside the /home directory tree.
This method allows me to omit things I don’t need. (Such as VMs)
There are a multitude of ways to automate backups with apps. You can even clone the entire drive if you want to. But a simple copy of the /home tree is the easiest. Learn to copy with “cp” from the terminal and it runs significantly faster than using the GUI file manager.
Timeshift can be used on fedora; but the directory names need to be renamed slightly to work with the software.
Side note: I hope your first foray into fedora is on an Intel mac (or a PC). Apple silicon is possible; but I would not recommend it for beginners.
I used Ubuntu for many years on a 2010ish MacBook Pro. It went from a death by bloatware to a better than new experience. I recommend you make this leap.
You will find that settling on a desktop environment will be more difficult than learning a few CLI techniques.
(Sorry for being off topic )
Maybe consider Pika Backup (website), it is basically a simplyfied GUI-wrapper around Borg Backup which might get you started.
Note that Pika Backup is deliberately NOT made for full system backups but rather for the important (personal) stuff.
It’s easily installable through the Gnome Software Portal.
I can understand that this may seem like a rather inconvenient solution compared to the one-click-restore-it-all approach as I had similar feelings for quite some time.
Upon closer inspection however, this approach gives you also the freedom to migrate rather easy to a different system-layout or filesystem or hardware or even Linux distribution at the cost of leaving the base-installation and configuration as a manual task to you.
Whatever route you take, it would be nice for the community to hear about your experiences and what you ended up with.
EDIT: Oh, a completely different approach for advanced users worth mentioning (no GUI, but making use of BTRFS’s snapshot functionality and thus closer to a full-system-backup solution): btrbk. I personally have never tried btrbk but it got on my wishlist for things to try out one day…
cp command does a full copy of everything while
rsync checks what is already there and only copies what is new or changed. A significant difference in the total amount of data written and time involved when doing large incremental backups.
My /home is over 4 TB and would take days to copy with cp (and require the data remain static while copying) while the incremental done weekly with rsync only takes a few minutes and is run overnight by a timer while the computer is idle anyway.
If you want the backup using rsync automated so you don’t need to use the command line it is very easy to write a simple one line script to perform the task then set that task to be run by cron or a systemd timer and forget about it.
Thanks but sorry I really didn’t understand much of that! Just copy the home folder?
I don’t use cloud anything.
You did say “you can even clone the entire drive if you want to” - THAT interests me. But I assume you’re talking about CLI with everything above, and whilst I will of course have to learn some over time after I switch, for now I need simple GUI ways to do certain things or I can’t make the leap. I am just so damn busy and that’s the reason I have tried and failed to do this many times over many years. I did run Fedora and Ubuntu on intel macs too, and agree it was a like new experience, but all the fiddling with wifi/broadcom/webcam… stuff, eventually I decided screw it. Sell all the hardware, get the right machine (thinking of a Dell XPS, I can’t stand the keyboard mushroom on thinkpads otherwise I’d have one of those), get Fedora installed hopefully everything working straight away, and get USING. That last part is where I have fears which prevent me forging ahead. I have hundreds of thousands of pictures and audio files, many tens of thousands of files, and I need to at least know my data is secure on Fedora (with backups or manual clones once a week will do)
Thanks. Pika looks well worth a look. I guess I have to learn what “my files” actually means. I have always enjoyed the comfort of knowing I have a bootable full clone, and that really is nice (for instance i can keep one in my safe with files going back 20-25 yrs, and go find one file if i need to rather than have all that trash clogging up my SSD on live machine).
I am starting to see what Kevin meant above about /Home. So that’s where all “my” data is stored, outside of system files etc. So my pictures, music, videos, spreadsheets and text files, email data etc etc?? If so, I think I am starting to understand, maybe a personal files backup would be more than enough to make me feel comfortable. Pika’s hourly backups therefore take me pretty close to TimeMachine experience. Thanks
That certainly sounds very cool! Still a bit ‘scary’ for me, but I can probably have a go at that. I guess I am just so damn used to a GUI app showing me my files being backed up, then reporting to me it has completed, and CLI feels scary because I don’t understand code at foundational level so I will be pasting in commands from online, and I never like doing that (i hate not understanding what’s going on fully, but have nowhere near the time needed to get that understanding!)
thanks everyone, some great suggestions and advice here.
Whether You use some GUI tool or rsync, You can see in any GUI file manager that Your files are backed up. And if You used Time machine on a Mac You shouldn’t be afraid of rsync. Allow me to quote JWZ:
Time Machine exists and is pretty good. It uses rsync underneath.
FWIW I have used Pika Backup for almost the entire time I’ve used Linux as my primary OS, and the same data / repository that I started backing up a year ago has carried through my experimentation with this PC - which involved installing or reinstalling the entire OS about 10 times in total, whether by force or by choice.
Family pictures, tax documents, the kids’ schoolwork, it’s all in there - backed up on a portable hard drive so we can grab and go if needed. (Over the past year I’ve figured out more things, so I’m also using rsync to sync those files to an old desktop PC in the house, and using rclone to sync to Proton Drive).
Interesting stuff thanks.
So rsync is a CLI command to synchronise (backup in this case) files from a set group of folders.
So the rclone command does the same to a cloud server, from CLI? Wow.
Yep - I have scripts set up now to take care of both of those, so the only thing I have to “remember” to do in the backup routine is plug in the portable backup HDD each day, and because Pika Backup is running in the background waiting to be able to run its backup job, then it automatically starts the backup within a minute after the drive is connected. Then I just check that it finished, unmount/unplug, tuck it away and I’m done.
And Proton Drive was incredibly painful to use until the 1.64 versions of rclone brought the ability to use Proton Drive as a remote connection - the Proton folks really owe the rclone folks something!
Wait, Proton Drive can be used on Linux via Rclone? Nice! I’m definitely looking into that.
Does it keep things in sync constantly like one would expect a sync program to do? Presumably this means Rclone would have to run as a daemon.
I intentionally don’t try to use it as a real-time sync, for me it’s a scheduled backup job - I imagine this page would have more info on that potential in general for rclone, but it does seem a bit dicey especially for something as slow and clunky as Proton: