0

I have a home server with the following specifications:

  • Lenovo TS140 Server
  • IBM ServeRAID M5015 RAID Controller
  • 4x WD Red 4TB Drives (WD40EFRX) in RAID 10
  • 3WARE SFF-8087 Cable (CBL-SFF8087OCF-05M) connecting the drives to the RAID controller

Recently I had a drive fail (after about 8 months of use) so I RMA'd it, received the replacement drive and rebuilt the array successfully. 24 hours later, the controller reported that the replacement drive failed, so I submitted another RMA. I just received the new replacement and, as soon as I insert the drive or try to rebuild the array, the controller's alarm goes off and shows the new replacement drive as failed.

The odd thing is that, if I take the drive out of the server and throw it into my desktop computer, I'm able to format and use the drive without any issues. Running a program which can read S.M.A.R.T. data (CrystalDiskInfo), the drive shows as being in "Good" health.

Unfortunately I don't have too much experience with RAID, so I'm not entirely sure what the issue is here. Should I just try sending the drive back for another RMA? Could it be the SFF-8087 cable that's failing? Or is there anything else that I can do to get a bit more insight into what might be causing this issue?

Here's a screenshot showing the drive's SMART data:

CrystalDiskInfo

I also ran Disk Self Tests (short and extended) in PassMark DiskCheckup and Western Digital Data LifeGuard Diagnostics. All tests passed:

WD Diagnostics Test Results

  • 1
    Sounds like a raid controller issue. – Moab Aug 8 '16 at 23:16
  • RAID controllers often mark their drives with some sort of a header that can be rather specific to a designer/manufacturer of RAID circuitry. If the drive works perfectly fine in another system that isn't using RAID, that indicates there is no RAID header to confuse the other system that isn't using RAID. The lack of header is likely what is causing the RAID controller to be unhappy. Not an indication of any bad hardware. You just gotta figure out how to tell your RAID controller to initialize the drive. Procedures vary by hardware; check documentation. – TOOGAM Aug 9 '16 at 1:09
1

Different controllers have different criteria regarding if a drive is healthy enough for it or not. Some controllers even refuse to get a new drive unless you rebuild the whole array.

Run WD Data Lifeguard Diagnostic and see if the drive passes both the quick and the extended tests. You may even do this with all of the drives. This should show you if the drives are healthy or not. You may also try filling the new replacement WD Red with zeros (full fill) with the tool and see if that fixed the issue with the controller.

Check the raw values of the S.M.A.R.T. status as they can give you more precise information about the condition of the drive.

Could you post a screenshot of the test that you did?

  • Updated the post with some test results/SMART data readings. I'm running the WD tool's extended test now and will report back with the results later. I'll also definitely try the WD tool's zero-fill capability once the test completes. – n110 Aug 9 '16 at 12:18
  • Everything looks fine on the S.M.A.R.T. status and the WD DLG shows the drive as completely healthy so I wouldn't imagine that the problem is coming from the drive itself. The only thing that catches my eye is the Spin-Up Time on the S.M.A.R.T. status. It shows that it takes roughly 7-8 seconds for the drive to reach it's regular spinning speed. This parameter is considered informational by the most hardware vendors. Although degradation of this parameter can be an indicator of drive aging and/or potential electro-mechanical problems, it does not directly indicate imminent drive failure. – Captain_WD Aug 10 '16 at 11:56
  • If your controller has a different approach towards this specific value and has a limit after which it considers a drive problematic (for example if a drive takes more than 6 seconds to spin up and this one does it in 7 seconds). – Captain_WD Aug 10 '16 at 11:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.