All storage systems fail, so be ready to recover
When saving data, don’t put all your faith in mortal hardware
- By William Jackson
- May 29, 2009
Any number of bad things can happen to a computer drive: fire, water, buses, cars.
The data on those drives can usually be recovered, said Michael Hall, chief information security officer at DriveSavers Data Recovery. But some situations are more critical than others.
“Where we see people really get into trouble is when they put blind faith” into a Redundant Array of Independent Disks, Hall said.
RAID setups used for data storage typically offer large capacity and a level of redundancy to improve the odds of recovering data. But redundancy is not immortality.
“The RAID is an electro-mechanical device,” Hall said. “It’s not a question of if it will fail, but when. It’s like a person. It has a life expectancy,” expressed in mean time to failure. A particular system might beat the averages, but eventually it will die.
However, there are ways to protect yourself from a loss of essential data. “The rule of thumb is always have a backup of the dataset on an alternate medium that you verify regularly,” Hall said.
Backing up data to a different medium from your primary storage helps ensure that a single incident does not damage both systems. And because it is easy for a backup system to miss some types of data, such as files that are in use at the time of a backup, it is essential to routinely verify that needed data is available in a usable form by conducting test restores.
“For the most part, people do,” Hall said, particularly organizations with continuity-of-operations and disaster recovery plans. “Everyone should have that.”
For those who don’t, there are professional data recovery shops to help you pick up the pieces.
William Jackson is a senior writer of GCN and the author of the CyberEye blog.