Server Crash Recovery Case Study: Lightning Strikes Twice

The client in this data recovery came to us after their local fire department suffered a server crash. A lightning strike to the station had led to a power surge. Afterward, the little two-drive server the department relied on to store their records became unresponsive. Both of the hard drives in the server’s RAID-1 array were clicking. The power surge had taken out both of them. The client sent the two Western Digital hard drives to the data recovery experts at Gillware to get their data back.

---

---

RAID-1 Data Recovery: When Redundancy Fails

The purpose of RAID-1 is to provide redundancy. While other RAID levels will often string multiple hard drives together to form volumes with larger storage space, RAID-1 isn’t too concerned with extra space. When you combine two drives into a RAID-1 array, both drives become perfect mirrors of each other. Anything you do to one drive happens to the other as well. This includes any data you write to the drives, anything you delete from the drives, or any changes you make to the partitioning scheme.

The idea of RAID-1 is appealing. If one hard drive has an unfortunate accident or dies peacefully in its sleep, the other drive has an exact copy of the data. Nothing of value gets lost. You can sleep soundly at night knowing that a whole half of your server is working tirelessly to make sure all of your data is backed up.

But in practice, RAID-1’s redundancy isn’t quite the security blanket it might seem. Since both hard drives see roughly equal amounts of use, they can have similar lifespans. Both hard drives can simply die of old age at around the same time. The second drive might die before the first can be replaced. Both hard drives are also roughly equally at risk for the same environmental traumas. In this case, a single lightning strike caused a server crash by damaging both hard drives.

Recovering from a Server Crash, Thanks to Gillware

This fire department’s RAID-1 server had gotten fried by the lightning strike. Both of the Western Digital hard drives inside it came out clicking. Power surges and power outages can damage hard drives in a lot of different ways. A sudden excess of power can fry the printed control board and possibly cause damage to the internal components of the drive. A sudden loss of power can lead to the read/write heads falling onto the hard disk platters and sticking to them, damaging both heads and platters. Power losses and electrical surges can lead to firmware corruption as well, and other fun things that turn your hard drive into a brick.

Clicking is often a sign that a hard drive’s read/write heads have failed. Clicking hard drives almost always need cleanroom work in order to function again. Our engineers opened up both of the client’s hard drives in our cleanroom data recovery lab to assess the damage. There was a lot of it.

Both of the hard drives’ sets of hard disk platters had sustained damage. The read/write heads in both drives had collided with the platters, scraping and scratching their surfaces. Failed read/write heads can create visible circular scratches on the platter surfaces. Severe platter scratching, or rotational scoring, can make a hard drive utterly unrecoverable. The material that’s been scraped off of the platters—where your data lives—can’t be restored. If critical areas of the platters, such as the firmware zone, get wiped out, data recovery becomes impossible.

One hard drive had suffered much more damage than the other. Its platters were severely scored on all surfaces, making data recovery impossible. The other hard drive, however, was in much better shape, with only a few light scratches on its platters.

Server Crash Recovery Results

When platters get scratched, our cleanroom data recovery technicians don’t give up. With the help of our platter burnishing technology, our engineers can polish the platter surfaces, cleaning the dust and debris resulting from platter scratches. After burnishing the platters from the one surviving drive and replacing its mangled read/write heads, our engineers could start recovering the data from it.

Our cleanroom technicians’ data recovery efforts proved a rousing success. Although one of the two hard drives had been scored beyond all hope of recovery, we successfully recovered 99.9% of the files from the second hard drive. Because both hard drives had failed at around the same time, only the one hard drive mattered. Our engineers ranked this RAID-1 data recovery case a 9 on our ten-point scale.