Scrivener & Backing up via an NAS

Greetings, so I spilled a cup of coffee on my macbook Air - luckily it’s working, so far haha. However, it was a bit of a wake-up call that I really need to back up my computer, and especially my Scrivener files.

So, I’m looking into NAS devices to back up all of my key content; and I’m wondering how best to do this with Scrivener - if I just copy the files do they maintain the file structures, etc ? How does this work exactly?

Also, can anyone recommend a NAS - they more or less all seem uniformly panned by reviewers on Amazon.

Love the new Scrivener update; I’ve been using this fabulous software since the beginning. 8)

I’ll let those with more tech knowledge advise you on a NAS. But a very simple, easy to implement “off-site” backup process is to sign up with a cloud service (google, OneDrive, iCloud, DropBox, etc., it really doesn’t matter) and set Scrivener preferences to save your zipped backups to the cloud service’s folder on your Mac/PC.

I use this approach, as do many on these forums. Let me know if you have any specific questions on how to set it up.

Once you get your hardware situation set up, use Time Machine. It’s simple, it backs up everything on your Mac, and it does incremental backups until you run out of space.

If you want to just back up scrivener projects, then make use of Google Drive, or some other cloud sync solution that has at least a 30 day retention of deleted files. You can then just set Scrivener->Preferences->Backup settings to point to that sync folder, and you’ll always have a co-located set of Scriv backups with minimal fuss.

Thanks guys. One thing I don’t really understand, though, and sorry if I’m being a bit dense: if I am saving/backing up/archiving Scrivener files in, say, iCloud - or, on an external disk - and the Scrivener software doesn’t reside in either backup place, are the files still the same? Are the projects (either in iCloud or the external device/NAS) still organized the way they are in my laptop copies? I know people say that there are no dumb questions but they haven’t met mine yet, lol :blush:


Don’t confuse Save with Backup.

When you create a project you specify where the project will be Saved, where you have the live project you are working on. Every time you stop writing for a few seconds, Scrivener will Save the project so it is safe if something happens (e.g the computer stops responding). So your current version of the project is always safe and saved. You can save the current project in a Dropbox folder so that it exists both on your computer hard drive and on the Dropbox server.

By default a Backup is created when you close the project or Scrivener, or hit alt-S to make an extra manual Save. Those backups should always be located somewhere else, away from your projects. If you save them as zip files, it is only one single file that is saved and you can then save it wherever you want, e.g. in a cloud service like iCloud drive.

Personally I wouldn’t use DropBox, Box, Google Drive etc for backups. They’re a sync service and if something happens to your hard drive and any changes corrupt the file you could finish up with two copies of a corrupted file. Yes you can do roll backs for a limited time but since it happened to me on one occasion I’ve always been wary.

I generally have my working project saved in DropBox with my backups elsewhere.

I have used a NAS drive - Synology 216+II. A useful device, the only things I had problems with are permissions which can be frustrating, and the fact that OSX is not case-sensitive or at least not as standard, whereas the Linux based DS boxes are. Most of the time you will get a warning but it can sometimes cause frustrations.

I have an Apple Airport with a USB socket into which you can push a flash drive or connect a 2.5" HD. I’ve sometimes used this for backups; load the drive on the desktop and drag and drop the backups to there. Finder -> connect to server -> server name (smb://Server_name) Some other modems do the same - Draytek I think and possibly some of the Billion range.

Of course if the house burns down… :smiley:

Scrivener projects live outside of the application. They’re just files, and don’t require Scrivener be on the computer/disk with them to exist.

Scrivener backups (as created via the Preferences settings I referenced earlier), are just copies of your project, often with some added information added to that copy’s name, such as the current date. There’s an option to zip them up (Macs hide the fact that your project is actually a folder with lots and LOTS of files inside), so that it’s easier to get a copy of that backup. If you’ve ever used “Save As” to create a copy of a Word document before you started making big changes to it, the principal is the same.

I went into Scrivener backups in great detail here: … s-backups/

I think it all still applies to version 3, though the screen shots will look a bit different from what you see on your computer. I hope that helps a bit.

But really, I recommend that you just use Time Machine with some kind of external file storage. It’s FAR better than no backups at all, it’s simple to set up, and it captures everything on your computer. Buying and setting up a NAS shouldn’t be necessary (I’ve been using the same external USB3 hard drive for years as my Time Machine backup), unless you’re keen on it.

To other posters;please don’t discourage the “better than nothing” solution of cloud syncing services. Not everyone is willing to go to the effort of the ideal setup. If we can get more people to occasionally plug in a usb3 hard drive for use with Time Machine + sending automatic backups to Google Drive, then those people are a 1000% better off than advocating for a backup setup that they’ll just give up on as too complicated. I know that I have given up on safer, co-located solutions for myself and my wife, and I work in IT–we even just moved our datacenter!

I’ve helped too many people who’ve lost data with cloud services to entirely agree with this. The speed with which a corrupted copy on one cloud-connected system will overwrite all other copies is quite impressive, and reverting a complex object like a Scrivener project is far from easy, no matter what the marketing materials say.

If you do choose to use a cloud service for backup, ALWAYS keep a local backup as well, in a location that your cloud service can’t touch. I would also recommend using a DIFFERENT service from the one you use for sharing live work between devices.

Whatever setup you use, TEST it. Make sure you understand where your data is going, and how to get it back, BEFORE your main computer gives up the ghost and you’re trying to recover your work on a strange computer with a tight deadline.


1 Like

Can this issue be avoided by using Time Machine → NAS (Western Digital for example)? Or would it present the same risk of overwriting with corrupted files?

But if a zip-file gets corrupted so the computer can’t read it, how could the computer then upload it to a cloud service?

Another method of backing up a Scrivener project (or any important document) is to zip it up and email it to yourself.

I have an email service from the provider of my .org domain, and it has kept some emails for me for over a decade now (I never auto-expire messages). Along with POP/SMTP, the provider allows HTTPS-based access to it, so I can get the old emails from any machine with internet access and a browser.

I also use daily USB(SSD) and NAS backups (RAID-5), but I don’t use a Cloud service. As an attorney, there are potential issues with storing client information on the Cloud that make it not worth it for me.

1 Like

I recommend that all Mac users have a Time Machine volume. In my experience, it’s much easier to go back to earlier versions with Time Machine than with Dropbox, and it stores enough history to take you back before the error occurred.


1 Like

There are lots of ways to corrupt a Scrivener project that do NOT render it unreadable by a computer.


That’s not an answer to my question:

Can a zip-file be corrupted in such a way that it can’t be read and its content can’t be restored but it can still be read and uploaded to a cloud service? I think not.

If the project is damaged when the ZIP file is created, it will still be damaged when the file is un-zipped.

You and I are not contradicting each other, but you are talking about ZIP files and I am talking about projects. The ZIP algorithm will happily compress whatever folder you give it, whether that folder contains a usable Scrivener project or not.

If something wipes out your entire Dropbox account, your ZIP files will be gone, too.

I agree with your point that a ZIP file is less vulnerable to synchronization errors or other damage in transit. But that’s orthogonal to my point, that cloud services alone are not a robust backup solution.


This actually happened to me just before the holiday. I was working at home with an approaching deadline, did a decent amount of work on the file and bang, a sudden power cut whilst I’d got an open project. Power outages aren’t very common around here so I’ ve never bothered with a UPS. However once the power came back I realised that the project was broken and corrupted and then the penny dropped that DropBox had been syncing in the background whilst I was trying to sort things out. Needless to say I now had two corrupted files.

Fortunately I had a backup elsewhere. One was on the Airport Disk, the other was on the hard drive which does have a Time Machine backup. A lot of faffing around and I had a workable project but I’d lost enough work to be a hindrance.

Could have been much worse and I’ve always taken the view you can’t have too many backups, but spread them around as long as you know where you’re putting them.

I learnt that lesson some years ago from a final year student approaching his project. He used Zip disks (yes it was some time ago) and religiously backed up to another Zip disk. All his fieldwork recordings and project notes were there all religiously backed up. What could possibly go wrong?

He kept both disks safe in his bag until one night the bag got nicked whilst he was in the Union bar…

That makes sense, thank you Katherine!

This might sound like a weird question: but how would I know if my files were corrupted?

Maybe I misunderstand, but strictly speaking, I don’t see why that would make a difference to the server. It copies your bytes even if they represent a corrupted zip string. I’m not aware of any service that would actively scan your files and reject them if they don’t open.

Your writing would be gibberish, the structure would be all over the place and your research documents would be meaningless.
No wait… that’s just my regular project. Sorry. :blush:

If you open a project up and massive amounts of data is missing, that’s usually a pretty good sign.