4 Disk - RAID(5) - 1 alleged faulty drive

Question

I have a OWC Mercury Pro Elite Qx2. I have 4 4TB Hitachi drives setup on a RAID 5.

One day, the red light came on indicating that my C drive has failed. I took it out, and put it on my docking station. Ran some diagnostics using OSX Disk Utility and SMART Utility. Reformatted the drive, the drive passed both the verification and the repair.

When I stuck it back in the OWC box, the "rebuilding array" light came on, a minute later it gave me the same red light indicating a failure. I am not sure why it is saying the drive is faulty. I have tested this drive pretty thoroughly outside of the RAID set and it seems fine.

All four drives are of the same model & firmware. They were bought at the same time, so its probable they came in the same batch.

Any thoughts on why this is happening? I am hesitant to purchase a new 4TB drive just to have the RAID go funky with putting in another drive of a different firmware/model.

Any help is appreciated.

Answer 1

TLER? Raid controllers hate drives without - once the drive has an issue. A client OS will live with the disc being stuck for a minute.

Answer 2

The only specific risk I am aware of from using a different brand for the new drive is if it turns out not to be exactly the same size, but rather be slightly smaller than the current drives. Using a bigger drive should be no problem, but may be pretty hard to find.

Your biggest risk right now is if another drive in the RAID fails. The more you delay the replacement, the greater the risk of data loss is.

Formatting the drive you removed from the RAID was not a clever move. If you have another drive failure, you'll be faced with a data recovery task, which would have been a lot easier, if you had not formatted that drive.

Taking the drive out, reading the SMART status from the drive, and reading the data from the drive sector by sector, would not be such a bad idea though. The SMART status could give a hint why the drive is marked as bad by the RAID.

For higher reliability I would recommend RAID6 or using a hot spare. Adding a hot spare is an easier transition for an existing RAID5, but it does not offer the same protection against data-loss as RAID6 does.

Answer 3

It could be that the controller recognized the disk's serial number and compared it to it's error log. Or, that it scanned SMART status of the disk and came back with an imminent failure. In any case, when a drive dies from a RAID array, it is normally recommended just to replace it with a new one of the same brand and model.