Everything fails eventually.

You would be shocked at how many friends, relatives and customers come to me with the same complaint: "My hard drive died, did I lose everything?" I've seen the entire range of emotions, and working through grief is a good thing. The sad thing is, these problems are completely preventable.

* Here's the first lesson, even if you don't read any further: hard drives fail.

I could expand that to say that everything fails eventually: your car won't drive for 100 years without a complete overhaul of most functional parts - making it virtually a completely different automobile; your home appliances will need to be replaced after a certain period of time; and each of your computer components will fail given enough time.

The term mean time between failures (MTBF) refers to a guess (a prediction) of how long a specific component will run until it is likely to fail.

Your computer has many working/functional parts: motherboard, memory, CPU, power supply and hard drive just to name a few. Each of these components has a MTBF associated with it, and this number differs between component and manufacturer. It's important to remember that even though a component might have a MTBF of - say - three years, it may fail after five years, and it may fail after a week of use - or even be non-functional out of the box (dead on arrival or D.O.A.). MTBF is an estimate, and sometimes these estimates are fairly accurate when you look at a large group of similar items: such as a million hard drives being used in the field. For your specific components, the MTBF is simply a guideline.

I mention the MTBF because there is often a correlation between this number and product warranty. Is it more likely to fail after a year? Then you are probably going to see a one year warranty. Do most of these items run for three years without a problem? You might see a two or three year warranty. You are not likely to see a three year warranty on a product with a one year MTBF, because that manufacturer would have a large number of returns, and they would be driven out of business.

What does this mean for prevention? It's simple: when shopping for components or systems, compare warranties. If you see a hard drive that has a one year warranty from one manufacturer, and a similar hard drive that has a three or five year warranty from another, there's a decent chance that the longer warranty translates to a longer MTBF. It doesn't always work the other way, though: sometimes a lower warranty is simply a manufacturer transferring risk to you, but the MTBF supports a longer warranty; just a business decision. You will rarely see things going the other way: a higher warranty means that the manufacturer (that wants to stay in business long-term) has mathematical, actuarial and historical evidence to support that warranty.

* Lesson two: when purchasing new equipment, look for items with the longer warranty.

Now there's a third easy lesson: try to replace components before they fail. How do you know when something will fail? You don't...not really. But you can guess. If a system has a three year warranty included, there's a good chance the manufacturer is betting that every component in that system is likely to work for at least three years. Knowing this, you can purchase a brand-new hard drive - perhaps a larger and faster hard drive - before the warranty is up. Have a computer person pull your existing hard drive, reinstall the operating system (Windows, Mac OS X, Linux) on the new hard drive, then place your original hard drive in an external enclosure - allowing you to copy your data from the old hard drive to the new hard drive at your leisure.

Replacing storage components before they fail is a great preventative measure you can take, and really not that expensive. For your average commodity laptop or desktop hard drive, you can often double what you currently have for around $100 - sometimes significantly less. Shop around the holidays, look for specials, and replace these components when they are on-sale. Just remember to purchase well-known brand-name, storage components and read the previous reviews on sites like Amazon and Newegg to see what other people have experienced with that specific product. Generally, items on-sale have been around for awhile, so you should be able to get a good sampling of reviews and ratings for any item you might be looking to buy. Again, check the warranties, as these are decent indicators of how often the item fails.

Lastly, and perhaps most importantly, everything I've talked about before have been proactive measures: buying components with decent warranties (and higher MTBF numbers) and replacing these items before they are likely to fail. However, if everything fails, and sometimes items fail significantly sooner than expected, what else can you do to prevent catastrophic data loss?

* Lesson four: backup.

Always have a backup of your data, applications and operating system. True, for any application or operating system you own, you should have a copy of, or access to a copy of, the installation media either physical or on-line.

* Side note: For systems that are pre-built that don't ship with physical media (a CD or a DVD for restore), your primary restore option is on a hidden part of your hard drive. If your hard drive fails, guess what happens to your restore option: it's gone. Call your computer manufacturer (HP, Dell, etc.) and ask for their method of creating physical restore media, or purchasing a copy of your OEM's (Original Equipment Manufacturer's) media for both the operating system (Windows) and any bundled applications like Microsoft Office.

Rebuilding a system isn't that big of a deal if you have all of the original components, but the one thing you can't reproduce without a backup is your data.

Years of music you've purchased and collected, photographs you've taken, spreadsheets and other documents; all of these could be lost forever - in an instant - if the hard drive where you are storing your data decided to choose this moment to fail. Lost forever, that is, unless you have a backup. True, there are data recovery options, but these can be very expensive - sometimes running in the thousands of dollars - and not always guaranteed to recover your data. What are guaranteed if you create and verify them regularly? Backups.

I will talk about backups in my next blog post, but suffice it to say, all of the planning and proactive replacement mentioned earlier can be relieved by taking regular backups of your data and, to a lesser degree, your applications and operating system. High-quality automated backup software can be fairly inexpensive, and the external hard drives needed to store the information are also relatively cheap.

Stay tuned...



Sean G.

Decades of computer and photography experience; patient teacher.

50+ hours
if (isMyPost) { }