Not directly a photography related post per se, but given the use of technology is now a days so embedded in most people’s photographic practice I thought I would jot down my thoughts on backing up and how I go about it. Hopefully I will keep it jargon free…
Your data, your photographs are important. Gone are the days when you had a physical copy of your picture – a printed photograph and a back up – the negatives. So if anything went wrong, you could always use your negatives to recreate the originals. Additionally your photographs may well have been stored in archival photo albums (as opposed to the big box on top of your wardrobe 😉 ) Either way, on the whole, your photographs were safe as long as they are kept in a dry place.
Skip forward a decade or so and your photographs are stored on your computer. Even if, as I was until recently, you are a film photographer you will tend to scan your images in to your computer having taken your photographs on your film camera. In your computer your data is stored usually on hard disks drives (HDD). An HDD consists of a spindle that holds flat circular disks. The disks are made usually aluminium, glass, or ceramic and are coated with a layer of magnetic material thats typically 10–20nm in depth. By way of example, a standard piece of paper is usually about 0.18 millimeters (180,000 nm) in thickness. Modern hard drives spin at either 5400rpm or 7400rpm – which is quite fast, but coupled with the fact that the heads that read the data “float” only tens of nm above the surface of the disk the risk of failure is there. The analogy about disks – whether it is true or not is another thing – is that if you were consider a jumbo jet flying 2.5cm off the ground the head of a disk drive when scaled down is closer to the surface of the disk.
Disk drives fail – in the past 10 years I’ve had 4 fail on me. All it takes is one to fail on you, and you’ll never consider it a waste of time to have a back up strategy. When my drive failed, I lost about 3 months of photographs – probably a good thing I hear some of you cry!! But it was a wake up call for me. By the time I had my second hard disk fail I had a back up solution in place – a Western Digital 500Mb USB drive if memory serves. With that back up, within a day of the failed disk being replaced I was back up and running again and our family memories were safe. I’ll touch on the other two failed disks later.
Backing up your photography doesn’t have to be complicated. A single external drive that plugs in to your computer via its USB port is a very good starting place. Most external drives will come with software that will allow you to select what folders on your PC that you want to back up to your external drive. The software will do this in the back ground and it will back the files up automatically.
Schrödinger’s backup “The condition of any backup is unknown until a restore is attempted.”
For most people in life, I think this will suffice. Its a relatively low cost solution to ensuring that your memories are on the whole a lot safer than they were if you were to continue to do nothing. However, let me introduce to the concept of Schrödinger’s backup “The condition of any backup is unknown until a restore is attempted.” You might be happily backing up your data, completely unaware that its valueless. Your back up hard drive might be corrupted. Your data that you’re backing up might be corrupted. You don’t know unless you have the luxury of time to practice recovering your files or if you have to do it in anger. How do you mitigate against that? Let me introduce you to the level of geekery that I partake in…
Defence in depth
I use various technologies to TRY and ensure that I have at least one good back up – I have over 10 years of photographs scanned and digital “originals” from my first forays in to film, to digital, to my wedding and the birth of my children. I also have the bad habit of never deleting anything (unless it is truly awful). As a result I have somewhere in the region of about 3 Terrbytes of photographs. That’s a fair chunk of data to protect.
First line defence
On my desktop computer I store all my data (ie photographs) on mirrored drives. Mirroring is as it suggests – I have 2 hard drives and using mirroring one is a as near as damn it real time copy (mirror) of the other. If one of these drives fails, the computer will still be able to access my photographs as it can still use the other drive. (If you want to be an ultra geek buy the same hard drives, but buy them from different resellers/shops to ensure they are from different production run batches, hence avoiding a “bad run”).
Rolling back the defences
So that is my “in machine” data protection covered. Stepping out of that and on to my desk, I have a Drobo DAS (Direct Attached Storage). This is connected to my PC via USB. The DAS (like a NAS – Network Attached Storage) is a essentially a box with a number of hard disk drives in it. These disks are configured using a technology called RAID – Redundant Array of Independent Disks. RAID 1 is disk mirroring that I described above. In the DAS I use RAID 4 – this means that any one of the 4 disks in the box can fail and my data will be safe. I can swap out the failed disk with a replacement and the disk will be rebuilt with the data from the other disks. You don’t need 4 disks for RAID 4- you can start with 3. The down side of RAID 4, harking back to my ultra geek comment, is that if more than one disks fails, you have lost your data and given that most shop bought DAS and NAS come with disks from the same batch, there is a risk that needs to be accepted or mitigated when considering your back up strategy. I have had two disks fail in NAS set ups before now – and its worked like a dream when replacing them and getting back up to “full strength”.
I use Memeo software for backing up my data – it takes a back up of any files in the folders I ask it to monitor. If you have the disk space you can also enable versioning which will allow you to keep multiple versions of the same file and “roll back” to a point in time. Memeo shouldn’t be used on any system drives and folders, as there is too much happening on those drives and it will be forever backing up. For systems drives where your operating systems and programmes reside, I carry out a full back up every month.
What I have described above is more than adequate for most people. However, I work in IT service management, and I have to expect the worst. When defining the requirements of data centres, I think of redundancy within the centre and then resilience across centres. So applying that to my little world of photography I think the worst and consider what if my house got broken in to and my computers were stolen along with any other kit….. I have a NAS in the garden shed that is doing exactly the same job as the DAS on the desk in my office. (I know, I know… 😉 )
Last Line of Defence (to the Cloud)
With the price of on line storage so cheap nowadays, it is now possible to store large amounts of data (photographs) on line. I use Amazon Glacier, which isn’t the slickest of services, but does what I need. Glacier doesn’t give you instant access to your files, instead there is a lag of about 4 hours between asking for a file and getting it. Which in my world is fine. For about $12 USD a month I get to store 3Tb in the cloud – where it is accessible to me. Uploading to Glacier is a manual process for me, I run a dif on what I’ve got on my machine versus what is in Glacier and upload the difference between the too – usually letting it run over night.
There are lots of other services that offer some kind of back up to the Cloud – Amazon Prime has a free version and there are other services that all automatically push your photographs to your preferred online storage provider.
Slicing and dicing
All of the above is just my take on things – your mileage may vary – as they say. You could have a PC and just back up straight to the cloud if you wanted to – if it works for you – then it is a perfectly acceptable solution. I would also add that a lot of my backing up is to do with geeking out a bit and trying different things out as opposed to arriving at the set up I did in planned for manner.
I’m always playing – I’m writing this looking at a 24Tb NAS that I’ve yet to configure. The idea is (I think) that with my local (eg house) network having Gb speeds. I may put all files on this NAS (and off the mirrored disks on the PC). This will allow me to access my files on the NAS over the internet (that’s more because I can than because I have to!).
So there you go. Hopefully I have kept the jargon to a minimum and I have made sense – apologies if I have not.
If you have any questions or comments, please drop me a line in the comments.