Intel 520 SSD causes system to lock up on format

[CHudk]

Hello!

I'm in charge of our company's fleet of managed assets. A few days ago, a problem surfaced with 240GB 520 series SSDs, utilized in Lenovo W510 laptops.

Summary: Upon clearing all partitions from the drive, creating a 2GiB and a 222GiB (approximate size), and attempting to format the larger one, after about a minute the entire system locks up, requiring a hard power-off.

Details

This is a scripted process, however for troubleshooting purposes I have also done it by hand, with the same result. The steps involved are:

1. PXE boot into Windows PE.
2. Remove all partitions (PowerShell Clear-Disk -RemoveData -RemoveOEM).
3. Initialize the disk (Initialize-Disk -PartitionStyle MBR)
4. Create a 2GiB partition and a second one using all remaining space (New-Partition ...)
5. Format both partitions to NTFS (Format-Volume -FileSystem NTFS -AllocationUnitSize 4096)
   1. 2GiB partition formats fine in just a few seconds.
   2. Big partition appears to work at first (HDD access light flickers as usual), but then something happens. HDD access light begins blinking regularly, and after about a minute the computer locks up hard.

Initially, I hit this with one drive and naturally thought it had failed. I tried with several other units, however, and swapped the drives between identical computers, also swapped them into different models, with the same result for four different SSDs. My script has not changed, and I verified that it still works on other hardware (using Crucial SSD in Lenovo W530).

I have run SSD Toolbox on three of these drives. Quick check shows both life expectancy bars at 100%, SMART shows no errors or warnings, and a long diagnostic turns up no problems. These drives have been in use for over two years, and most of them have about 4 TiB written according to SSD Toolbox.

I just tested with a SSD that had been through a detailed diagnostic with SSD Toolbox; I created partitions using the Windows Disk Manager, and formatted them just fine. I put it back into a machine and attempted to format using my script. Machine locked up.

At this point I am stumped.

Where would we go from here?

[CHudk]

Ah, I had not considered potential differences between OEM and retail versions of the same product.

From what I can see, the drives do have the latest firmware -- at least, the ones I have checked do. SSD Toolbox reported no updates available, and I can get the revision easily using this:

(Get-PhysicalDisk | where MediaType -eq "SSD").FirmwareVersion

LF1i

Do I understand correctly, that the "L" there denotes a Lenovo firmware?

Also, because you said you've found Lenovo discussions and advisories regarding this drive, I have been searching for them. I found a couple of forum threads (such as https://forums.lenovo.com/t5/ThinkPad-X-Series-Laptops/x230-SATA-errors-with-180GB-Intel-520-SSD-und... this, and https://forums.lenovo.com/t5/ThinkPad-T400-T500-and-newer-T/T430s-Intel-SSD-520-180GB-issue/td-p/888... this) but those concern the drive locking up under heavy load -- which is certainly not the same problem I'm encountering. (Indeed, the SSDs have all performed very well in that regard; their usage here has them basically setting up a new OS every week, and often running virtual machines on top of that.) Would you be able to link me to any of the discussions you found?

I ask because if this is a known issue with Lenovo, that would help me make a case for getting them to replace these drives.

[jbenavides]

The firmware version in fact indicates that these are Lenovo OEM drives, Intel® SSD Toolbox can't update the firmware on OEM drives, so it will show that no updates are available. It is worth to mention that the retail version of the Intel® SSD 520 series did not get any firmware updates.

The forums you found are the ones I saw before. If your computers still have support from Lenovo, you might be able to check with them on this issue as well.

The fact that you are able to initialize the drive normally from disk management suggests there may be a problem with the variables in the script, or a possible firmware issue.

- Have you tested if you have issues to partition and format the SSD's using other methods?

- Have you had success with any drive of this type?

[CHudk]

jonathan_intel wrote:

The fact that you are able to initialize the drive normally from disk management suggests there may be a problem with the variables in the script, or a possible firmware issue.

I agree. That's a reasonable idea, and (as I believe I mentioned earlier) I did check the script. I had not made any changes to it between the time when it worked properly (three months ago) and the time when it failed (five days ago). You mention that there have been no firmware updates for this device. So if...

1. Script has not changed.
2. Firmware has not changed.
3. BIOS has not changed.

...a reasonable conclusion would be that something about the SSD itself has changed. This could include obvious statistics such as hours powered on, bytes of data written, number of power cycles, etc. These would naturally increase with usage over time. Non-obvious things... well, I guess I haven't found them because they aren't obvious.

- Have you tested if you have issues to partition and format the SSD's using other methods?

- Have you had success with any drive of this type?

Yes, I've tried a couple of other methods:

• Scripted using PowerShell in Windows PE 5.0: Device locked up
• By hand using PowerShell in Windows PE 5.0: Device locked up
• By hand using DiskPart in Windows PE 5.0: Device locked up
• By hand using PowerShell in Server 2012 R2: Worked normally
• By hand using Disk Management in Server 2012 R2: Worked normally

As for "success with any drive of this type," I'm not sure I understand the question. These very same drives worked a few months ago with the same procedures. The same script also works with HDD-based machines, and also with a set using Crucial SSDs.

We use the same process to set up dozens of computers each week, and hundreds every month; all of the laptops in our fleet are refreshed quarterly, so I've done this a lot. Because of this, I have an interest in finding out why this is happening, so that I can determine whether it is likely to happen to any of our other hardware.

[ppara5]

Excuse me for butting in, but I may have a solution. Somewhere on intel.com (sorry, I did not retain the link) is a reference to DBAN. The Intel article suggested that if an SSD does not respond to the usual treatment, running a single-pass zero-write (the top option in DBAN's list) will completely wipe the drive and return it to a usable state. DBAN's website specifically states that it is not designed for SSDs, but I tried it. It allowed a drive which had screwy errors to one which was usable again. I have done this on Crucial and Intel SSDs, but it's only a last resort.

[CHudk]

paramountain wrote:

Excuse me for butting in, but I may have a solution. Somewhere on intel.com (sorry, I did not retain the link) is a reference to DBAN. The Intel article suggested that if an SSD does not respond to the usual treatment, running a single-pass zero-write (the top option in DBAN's list) will completely wipe the drive and return it to a usable state. DBAN's website specifically states that it is not designed for SSDs, but I tried it. It allowed a drive which had screwy errors to one which was usable again. I have done this on Crucial and Intel SSDs, but it's only a last resort.

That's fine, thanks for the suggestion! Actually the link above (in Jonathan's first reply to me) has a page with links to DBAN, KillDisk, and another program. I have already done the zero-write procedure using KillDisk and it made no difference.

I've also used DBAN before for wiping drives of work and personal machines, but in this case I chose KillDisk because I didn't want to make a boot disk.