Strategy, backup strategy.

I'm sure you have read those articles that explain how and why to back up your data. If you are somewhat interested in technology those articles are hard to miss. 

I think there are three groups of people when thinking of backing up computer data (your precious photos, your almost finished thesis or novel, that statistical data that was so hard to accumulate). 

The first group never makes a back up. They think it's too much hassle but know they will start making them in the near future, or they have no clue what a back up is. 

The second group does makes a regular backup, be it in the shape of Time Machine, some other software solution that takes care of it or through old fashioned human intervention. Most notably on an external disk.

The third group doesn't simply make regular back ups, but has a back up strategy in place. 

After some sort of data loss, the people in the first group often, but not always switch to the second group. Not always though.

After some sort of data loss the people in the second group often switch to the third group. Again, not always.

As a bit of a nerd and someone with a fear of letting go, I started off in the second group, but moved to the third group the moment I learned there even was such a thing as a ’back up strategy’.

So what is this ‘back up strategy’ all about?

Hard drives are mechanical precision machines, with very delicate moving parts. Normally a every hard drive will last around 3 years, maximum. 

This means that the external drive you use for a back up will die within about 3 years of use. That is to say if the disk was perfectly mechanically sound to begin with. The disk didn’t get knocked down, nor was it exposed to undue vibrations during it’s use.

Then there’s the fact that the magnetically charged areas, loose strength or due to some external event get overwritten. 

Someone who has a basic strategy in place, backs up to more than one external drive. So there is no single point of failure.

One way to mitigate the risk of a single disk failing is the use of a setup called RAID. This RAID system is usually a box with at least two disks installed. The box can be configured to make a backup of your system on each of the two, or more, disks in the box. That way you have two, or more, identical back ups. 

The RAID system can also be configured that the data it backs up is distributed over all the disks in the box, be it 2, 3, 4 or more. This way when one of the disks has a fatal mechanical breakdown, you simply replace it with a new identical disk, and the RAID system will rebuild the data and start using the new disk. No data lost.

A RAID box may seem like a good solution, but you still have the single point of failure: the RAID box. It too has a microprocessor and an operating system. Those can break or act up, without you even noticing at first, and suddenly you have no backups.

I’m not saying that a RAID system can’t be part of a backup strategy, it most certainly can, and I’m sure every large corporation has tons of RAID systems set up in their data stores, but in combination with other backup options in operation. 

Data loss experts like to say that you don’t actually have a back up if you don’t at ALSO have at least one back up OFF site. So when your office or home burns down, gets broken into, you still have the back up in that OTHER location. I’m certainly not a statistician, but the odds of both locations losing the back up is rather slim, I’d say.

So what made me go on about this back up strategy thing all out of nowhere?

The other day I noticed that a considerable number of image files were missing from the hard drive on my iMac. Really strange, I had never been faced with a situation like that. They had just disappeared.

I looked on the main back up drive, it backs up the entire system at midnight every night (it actually simply makes an exact copy of the hard drive, experts don’t call this a back up per se). Gone. The files weren’t there either. That meant that the files had gone missing before the back up was made. It’s an exact copy of the original so it would make sense that what’s missing on the original is also missing in the copy.

So I checked Time Machine, the built in back up facility of MacOS. Luckily I found an intact version of the folder structure where all my images reside. So I had Time Machine restore the images.

Only to discover, that they went missing again. 

I checked one of 5 external drives that I knew would, under normal circumstances have all the files. It did! So I made sure not to connect that disk to my iMac, I didn’t want whatever was making the files disappear erase them from that disk too.

In fact I suspended all automatic back up tasks. With all back up and remote synchronisation tasks suspended and all but one back up drive and Time Machine disconnected from the iMac, there was no foreseeable risk of losing data from the disconnected disks.

In the end Time Machine managed to restore all the files and after a few days there were no disappearances anymore. My files are safely back where they belong.

Did I panic when I noticed precious images from unique travels, like our journey through Bhutan, our encounter with Perito Moreno Glacier were gone? No. I knew I would have a copy of them. Either off site in Munich, or on one of the 5 disk that I use in my back up strategy.

I never figured out what caused the files to disappear. Which is not a very comforting thought, I have to admit. I suspect there was some sort of conflict with the synchronisation software I use to have an exact copy of my files off site.

I will now slowly reimplement automated tasks one by one, making sure each is preforming as needed. And I will keep one disk out of the loop until I’m sure things are running smoothly again. That way I will have at least one disk that was never affected by whatever caused the problem!