RAID Configuration: 4 Seagate Barracuda Green 2 TB SATA hard drives (model ST2000DL003) configured as a RAID 5 connected to an on-board Intel RAID controller.
Problem with RAID: Because of the nature of Green drives, specifically that they will spin down in low use periods, the array will periodically have a disk report as offline. The fix this customer was using was to mark the offline drive as active and rebuild the RAID. Unfortunately, a drive eventually did fail and on trying the rebuild it would not come back up.
The customer then proceeded to buy a similar hard drive, replacing the failed RAID hard disk and try the rebuild once again. This attempt to rebuild the raid was not successful either.
Our Advice: Always proceed with caution when dealing with a failed RAID array, especially if you are looking at performing a rebuild. There are many things that can go wrong during the rebuild of a RAID 5, some of which may make recovery attempts impossible. The rebuild could only partially finish, it could re-initialize the array causing full data loss, a stale drive could get rebuilt, or the drive you think is bad may not be labeled correctly or the position you think it is in may not be the position the RAID controller sees it in.
The RAID 5 failed for a reason. The fact that a drive would be marked offline frequently should have raised a red flag. At that point, the disks should have been replaced before the complete RAID failure. It is advisable to only use enterprise class hard drives that are designed for RAID 5 configurations and to check the hardware compatibility list for the RAID controller you are using. RAID controllers can be quite temperamental when it comes to disk responsiveness. A RAID disk with slow read times, or slow spin up times can fail continually.
Your best bet when faced with a RAID 5 failure is to contact someone with RAID recovery expertise.
Important RAID 5 Recovery Steps
- Write down the RAID 5 configuration and other information if available. (RAID type, stripe size, drive order, partitions, etc…)
- Pull each RAID disk from the array. Mark the hard drive and the cable it was connected to for order identification.
- If the Bad drive or drives are known, mark them as well.
- Hook each hard disk drive to a computer and clone to a disk of equal or larger size. See “How to slave a hard drive”.
- Use a sector cloner to image drives.
- Mark any drives that experience issues during the cloning process.
- Keep the hard drive clones and send the originals to a hard disk recovery company.