Backups are something that we’re all meant to do on a regular basis, but often don’t. Here I’ve collected how I currently do it, and what I think are the important things to remember when backing up. As always, any backup is better than nothing, so start if you haven’t, and then iterate and improve on it later.
Backup 3-2-1 Rule
There are slight variations and interpretations, but generally the idea is:
Three Copies of anything important (e.g. one on your main device, and 2 more)
Two media. We’re not talking about tape anymore, but you shouldn’t have two copies on the same hard drive or device. That might protect against corruption of one, but doesn’t against disk failure, loosing the device etc.
One off-site location. It shouldn’t be in the same place as all the other copies. This protects against things like fire, flooding and theft
The first two points are relatively easy to achieve using either a NAS or an external hard drive. In most cases your operating system comes with a built in way to back up automatically. MacOS has Time Machine, and Windows 10 has an inbuilt back up and restore system.
If you’re on Linux then using Rsync is probably the easiest option. You can write that into a script and run it on a regular basis with a cron job.
A NAS is normally easier to automate with as it’s always available. Unless you have a computer with a second built in hard drive, or one that’s always connected to an external drive. Otherwise you have to remember to connect the external drive before letting the automated back up run.
Accessible Backup
The most important part of a back up is the ability to restore from it. You might also want to access the contents if your original device is no longer functional. Therefore you should be prepared to be able to access it from a different device or operating system.
If you use something like Time Machine, you will need another Mac to access it, you can’t just open the files from say, a Windows computer. This way you won’t be forced to quickly buy another of the same type of computer, if you don’t want to.
On a NAS you don’t have to worry about the formatting of the hard drive so much, as there is normally some operating system layer in between that serves you the files. With an external hard drive this is also something to watch. Cross file system compatability is astoundingly poor even in the 21st century, and there’s not really a stand-out winner. You should check which best matches the mix of devices you have, but it’s probably going to be ExFAT or, surprisingly, NFTS for highest cross-compatibility. In many cases you can get away with read-only on your backup, from another type of system.
The Cloud
Currently ‘The Cloud’ is often considered the answer to almost any computing problem, and with plentiful storage located off site, it might seem the obvious answer to being a back-up solution. It has its place, but there are limitations.
First, storing something in the cloud means it’s much less private than storing it locally. ‘The cloud is just someone else’s computer’ goes the old1 saying. This means it’s easier to other people to access it. If that’s a concern, but you still want to use cloud storage then you have to make sure it’s encrypted, but that adds a layer of complexity - and if you need to access it from somewhere else, you need to make sure you have some way of decrypting your backup, so you have to keep a copy of the key safe somewhere else too.
Second, if you’re using cloud storage in the sync sense (e.g Dropbox, Google Drive or OneDrive) that mirrors the content of your local files to/from the cloud that’s not really a back up. If you accidentally delete something in one place it’s gone everywhere that these files are synchronised (or at least moved to the trash).
Lastly there is the cost. Cloud storage is becoming cheaper and cheaper, but it’s a cost you will have to pay every month. If you’re backing up a lot of data, the cost of a few month’s storage will probably be that of a USB hard drive of equal storage volume, and you only have to pay that once.
There are cheaper cloud storage options, something like AWS Glacier , but that’s not quite as easy to use.
W hile cloud storage has its limitations, there are also advantages. The main one being it’s off site, but in a way which is accessible from lots of different locations. If you’re travelling a lot, then you’ll want some kind of remotely accessible back up location.
That’s why for something like the photos you take on your phone, enabling cloud backup (often enabled automatically on Android or iPhones) makes total sense. You should still copy these somewhere else though.
My Setup
Given all of the above, what do I think makes most sense?
First, my main computer gets regularly backed up the NAS in the house, using rsync . That way there’s a back up that’s almost instantly ready and can be accessed from any other device I can log into the NAS with. I actually don’t have the back-up script on a cron job, I set it off by hand. This is because I split it into three parts of about equal size: music, photos and documents. I know music and photos don’t change as frequently, and when they do, I’ll run a backup. As my main computer is a laptop, it’s not on when I’m not using it, leaving a script to run every day at a time I’m not activly using the computer would never run. Instead I launch manually when I’ve finished working.
The NAS also contains the back ups of other devices in the house, plus other larger files that I don’t use regularly. This means it’s larger than the content of my laptop, and the most economical option is to Rsync the contents to an external USB drive on a regular basis.
I actually do this with two drives (that’s why them being cheap is important) in rotation, with one always being kept off site. I think this is a good alternative to using a cloud solution. As long as it’s somewhere you visit regularly and isn’t too far away, like a friend, relative or perhaps your workplace, then it’s a good choice.
The other advantage of this is that it means this third back up is air gapped. Should anything malicious get into my computer, or the data on the NAS, what’s on the unplugged hard drive cannot be affected. This is something, that perhaps incorrectly, worries me about a cloud back up via an always on synced connection. If it can be accidentally, or intentionally, deleted at the same time as your main copy, it’s not a real back up.
‘Old’ in the internet sense. ↩︎