The Importance of Backups

Operating systems and applications can be re-installed with relative ease, but personal data is just that, personal. Nobody else (hopefully) has any copies of it, so if you lose it, that’s it, it’s gone forever. For this reason, it’s important to keep backups of your personal data.

That being said, backups of reproducible data can still be very useful as well, if the time it would take you to recreate said data is more valuable to you than the data itself. For example, it’s very easy to get an operating system set up the way you like it, but if you want to get on with creating things, instead of setting things up, then it’s worth having backups that allow you to get going again quickly should the worst happen.

What should you backup? How often? Why?

As we touched on above, you should back up anything that is irreplaceable (photos, letters, nan’s recipes), and anything that would take a non-trivial amount of time to recreate. How often to backup your data depends on a few things:

  • How often is it changing? Taking daily backups is pointless if your data is only changing once a week. On the flip side, backing up once a week if you’re data is changing daily also leaves a lot of room for lost work, which brings us to our next point
  • Granularity – how much detail would you like your backups to cover? Lets use a novel you’re writing for an example. How often would you want to save copies of your work? Every page? Every paragraph? Every line? Every word? Whilst this isn’t the best example, seeing as storage is so cheap nowadays you could store every different letter and get away with it, it illustrates the concept nicely. Even if a paragraph in your novel only takes a few minutes to write, what about the ideas in that paragraph, can you guarantee that you’ll think of the same great words next time if you were forced to rewrite it? Make sure you can track changes in the right detail can make all the difference between having backups, and having useful backups.
  • Storage costs – Let’s scrap the novel idea for now and think big, really big. Take a media production house, they’re gonna be storing Gigabytes, maybe Terabytes of data per project. This can result in some serious costs for your storage hardware. Unlike the novel, the cost of saving a copy after every change would be prohibitive, so you need to draw the line somewhere else. Where this line falls again comes back to the time-cost comparison: how much will it cost you to store the backups, and how much would it cost you to carry out the work again? Missing an important deadline to lost data is a real pain and can make you look unprofessional.

The 3-2-1 Rule

This is a common rule when talking about backups, at least at the simpler levels. The rule dictates that you should always aim for:

  • At least three copies of your data
  • On at least two different storage mediums
  • With least one of these copies in an off-site location

For example, one copy of the data on on your hard disk, one copy on an external drive, and a final copy in the cloud. This gives you a great chance of recovering your data in the event of problems. With the ubiquity of moderately priced external storage, and the plethora of free cloud storage solutions out there, it makes it really, really easy to have multiple copies of your most valued bytes.

RAID, why it’s great, and why it’s not a backup

RAID (Redundant Array of Independent Disks) is a technology that used to be found solely in the enterprise. However, as with most things in the tech world, it has found it’s way down into the levels of your everyday users over the years. RAID allows you to keep multiple copies of your data automatically and transparently with relative easy. At the user level, you save you data just as you normally would. But behind the scenes, clever bits of hardware and/or software makes multiple copies of this data and store it on multiple physical disks. In it’s most basic form, RAID-1, also know as a mirror, does just that; mirrors your data. One file (or millions of them) stored identically on two (or more) disks. If one of the disks stops working, you can grab your data back from the other disks. Great right? Yes.

However, RAID is not the solution to all of your backup woes. RAID’s strength can also be seen as it’s downfall, and that is that it does things automatically. If you delete a file, it’s deleted from all of the disks. It’s not clever enough to realise that you didn’t actually want to remove that file forever. Remember, computers are dumb, they just do what you tell them. RAID can be seen as increasing the availability of your data. It saves you having to pull copies from your other storage methods from the 3-2-1 rule. What it doesn’t protect again is somebody clicking the wrong button and washing away all of your favourite pictures.

This is another advantage to the 3-2-1 rule. Even if you delete something on your primary storage, chances are that you realise before you sync this storage to your secondary storage. And if you don’t catch it then, then chances are you will catch it before you sync things to your off-site storage. These layers offer time delays allowing you to realise your mistakes and correct them.

Testing your backups

Testing your backups is of critical but often overlooked importance. Having all the backups in the world is still no good if they don’t actually work. For this reason, you should try to verify your backups are in good condition as often as possible. Make sure that novel opens fine in your text editor, make sure some of those family photos aren’t missing etc etc. It’ll be devastating to find out your golden backup solution is anything but in the times when you need it most.

Let us help

If in reading this blog post you’ve had a panic and realised your server is lacking any meaningful backup solutions, then please get in touch. We’d love to get your data stored away safely for you.

 

Feature image background by gothopotam licensed CC BY 2.0.

0 replies

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published. Required fields are marked *