Back in February, Google released a report titled » Failure Trends in a Large Disk Drive Population (13-page PDF, 242-KB), which detailed their findings regarding the failure rates of hard disk drives they own.
Since Google uses so many hard drives (over 100,000 were used in this study), their findings were of great interest to those of us who've had trouble locating reliable data on the topic.
Moreover, their report was of special significance because it contradicted many commonly held beliefs, such as:
- cooled drives fail less
- the harder a drive works, the more likely it is to fail
- SMART monitoring is a reliable predictor of drive failure
- Mean Time Before Failure (MTBF) is a meaningful spec
Personally, I think drive failure has a lot to do with how the drive was handled en route (by the shipper) » from the manufacturer to the reseller, and » from the reseller to the end user (you & me).
Hard drives aren't designed to be tossed around like fish at the market, or dropped like a hockey puck. And these are factors no one will ever know (except the carrier).
What the survey *did* confirm, however .. was something we already knew » hard drives fail more than we'd like.
The Google survey reported that ~2% of their drives failed during the first year of operation, and ~8% every year thereafter.
So how do you protect yourself against the devastating effects of hard drive failure? There are various ways, with each method offering a degree of protection. But the *best* way to protect your data and your operating system is with an imaging/cloning program .. such as Norton Ghost (distributed by Symantec).