I’m paranoid about data backups. We should all be really considering how much of our personal data is stored on our various devices. For me, it’s not just the precious data such as family photos, videos and such like, but of course also all of my business data and media that I create for clients. Having a robust, easy, automatic back-up strategy is therefore vital.
I’m not sure where I first heard this, but there is a 3-2-1 back-up strategy that it is recommended we try to adhere to. That is:
- 3 – At least 3 copies of any data. The original copy and two backups.
- 2 – At least 2 different types of storage.
- 1 – At least 1 offsite backup.
My strategy covers all of this and more.
My first back-up is a standard Apple Time Machine Back-up. This is a full back-up of my data but I don’t really treat it as a back-up more as a fail-safe that allows me to roll back in time should I change something that I would later like ‘un-changed’ or delete something that I’d later like undeleted. It’s saved me a few times over the years and is so easy to set up and run that everyone on Mac OS X should be doing it.
My next line of defence are cloned back-ups using Super Duper! These are exact clones of my internal hard drive stored on external drives. The idea is that should my internal drive fail catastrophically I should be able to restore my cloned back-up to a new hard drive and continue exactly where I left off. I can even boot from one of the cloned external drives in a pinch so that downtime should be minimal.
My main clone runs daily. The only issue with this is if I need to restore my hard drive due to a software issue such as a virus, malware etc. In this case, it’s likely that the daily cloned backup will have already backed-up the malicious software that is causing the problem. Because of this I also run weekly and monthly cloned back-ups so that I can roll-back a little further should the need arise. In theory, even if I haver to restore from a month old clone then I should subsequently be able to restore any newer files from one of the more recent clones as well (being careful to only restore files that I know are safe).
Super Duper! runs all of these cloned back-ups for me and does it on a schedule so that it is all automatic and I don’t need to do a thing. Although I do sometimes test the back-ups to make sure all is well.
Off site Back-ups
This is where it gets interesting. In the past I used to use a system called Carbonite that backed-up my data to a server in the ‘cloud’. There were a few issues with this, mainly related to our slow Broadband connection. The initial upload of all my data took about 6 months and even then, it had trouble keeping up with the upload of new data. Despite running it continuously there was often a backlog.
This was made worse when I bought a new computer. Even though I cloned the hard drive from my old computer to my new one, Carbonite couldn’t recognise it as the same data and therefore required me to re-upload all of my data. By then I had more data than before so the upload was going to take 9 months of continuous uploading. This was made worse by the fact that Carbonite keeps deleted / removed files for 30 days before removing them from your back-up. As it saw all my data as being new, it wouldn’t recognise that the files in my back-up hadn;t been removed from my computer and would therefore have deleted my entire back-up. I would have been without an off site back-up for several months until my new back-up had finally uploaded.
With this in mind and after lots of support calls to Carbonite to see if they could help I decided to cancel my Carbonite subscription and revert to the old fashione ‘sneaker-net’ method for offite back-ups. I bought two portable external hard drives, made cloned back-ups to them and then stored one of them in my office at the University. I then just swapped them over once a week. The simple solution to off-site back-up but with one small problem. I’d often forget to run and swap over the drives so the off-site back-up could become out of date.
Things have changed a little since then.
- I have even more data to back-up than before
- We now have faster Broadband thanks to our Fibre to the Property connection – with upload speeds in the 10Mbps range rather than the old 0.2Mbps range
- We have an Amazon Prime account and can therefore get unlimited Amazon Cloud Storage for £55 a year.
The faster broadband and access to unlimited cloud storage should mean that I can start with an automated Cloud back-up strategy once again for my off-site back-ups. I came across an app called Arq Back-up that I’m trialling, along with a 3 month trial of Amazon’s unlimited Cloud Storage.
Arq Back-up seems pretty good so far. It allows you to back up to your own Amazon Cloud Drive, Amazon Web Services, Google Cloud Storage, Google Drive, Dropbox, or OneDrive accounts, or your SFTP server or NAS. Arq Back-up encrypts your file data before it leaves your computer, so your files are never sent on the network in the clear. It uses various technologies to make sure that uploads are as efficient as possible:
- Compression reduces upload sizes.
- Client-side de-duplication ensures Arq never uploads the same content twice.
- Rsync-style “rolling checksum” ensures only the changed parts of files are uploaded.
Arq Back-up also has a nice Mac-like user interface and a one-time only fee for purchase rather than a monthly subscription. So far so good, but I am only testing it. Even with a new superfast BroadBand connection it’ll take a while to upload all of my files. I think it’ll take about 6 or 7 days of continuous uploading to get the intial upload into the cloud, but that’s MUCH better than the 10 months it would have taken with our slow Broadband connection.