Repeated, disappearing SMART warnings

[Prodigal]

I'm running the scanner on my server that's running WHS 2011. Every night I get an email about how one of my drives has 6 unstable sectors, and another drive has 132096 unstable sectors. One minute later, I get another email that only says I have one drive with 6 unstable sectors. If I log into the server and look at the dashboard it shows the latter drive has not been scanned.

 

This has been happening every day for the better part of two weeks. Last night it happened 3 times in 3 hours.

 

The drive that has 132096 unstable sectors (apparently for only a minute at a time) is a brand new SSD that I installed ~2 or so weeks ago. The server itself is running fine, no noticeable issues.

 

Today I noticed that I must have disabled scanning or something like that at some point because the "Start/stop automatically" button in the dashboard wasn't pushed. So I pushed it, and it immediately starts scanning the SSD and an external drive that it hadn't previously finished. About halfway through the scan I get another automated email about the 2 drives with unstable sectors (identical to all the previous ones), and 2 minutes later another email about the one drive with 6 unstable sectors only.

 

In the dashboard the SSD now shows as Healthy, and unstable sectors isn't even a category in the SMART details for it. Another odd thing I just noticed is that it reports its temp in one of the SMART categories on the same detail window, but reads as 0 both at the top of said window, and in the dashboard.

 

Any ideas on what's going on here?

[Christopher (Drashna)]

You need to install this version: 

http://dl.covecube.com/ScannerWhs2/beta/download/StableBit.Scanner_2.5.2.3121_BETA.wssx

 

There is an issue with disk identification (especially on newer disks), that can cause this sort of issue. 

 

 

If the issue persists after installing this version, let me know. 

[Prodigal]

So I installed that beta version, and it proceeded to rescan everything. I told it to ignore the 6 unstable sectors issue unless it gets worse, and all drives were reporting healthy.

 

Then at around 1am (same time window as before) I get an email that the SSD has 132096 unstable sectors again. When I log into the server and pull up the dashboard this morning, it reports the SSD as healthy.

[Christopher (Drashna)]

So I installed that beta version, and it proceeded to rescan everything. I told it to ignore the 6 unstable sectors issue unless it gets worse, and all drives were reporting healthy.

 

Then at around 1am (same time window as before) I get an email that the SSD has 132096 unstable sectors again. When I log into the server and pull up the dashboard this morning, it reports the SSD as healthy.

 

To make sure, this is on the 2.5.2.3121 build, correct? 

 

If so, could you run a burst test on the disk (right click on it and select the option). 

 

Also, could you grab the "System" section of the Event Viewer (run "eventvwr.msc), and upload that?

[Prodigal]

It is indeed running 2.5.3121 according to the Add-ins section of the dashboard.

 

Burst test has run for the last 12 hours successfully w/no errors that I can see (and is still going).

 

Here are the event logs for the server for the last week: https://www.dropbox.com/s/wsek9aw4qsjt44u/eventlog.evtx?dl=0(I get an error about not being allowed to upload that file type when I try to put it on the forum itself).

 

Also I received the unstable sectors email twice since starting the burst test. Once within a few minutes, and the second ~11 hours later.

 

• ADATA SP550 ATA Device - 1 warnings
  • These are currently 132096 unstable sectors on the hard disk. An unstable sector is a sector that can't be read. The drive will automatically swap the bad sector for a good one whenever new data is written to it, however, the original data may be lost.

 

[Christopher (Drashna)]

Could you open the Scanner settings and verify (it should be at the bottom of the Window). 

Sometimes the dashboard is "wrong", as the data is maintained independently of the installer info or actual files.  

 

And the burst test runs with no issues, but you're still seeing it flag unreadable sectors on the disk? 

 

If so, what controller is the drive connected to? 
If it's an intel controller, do you have the Intel RST drivers installed? 

 

 

And I'd recommend running "chkdsk x: /b", where "x:" is the disk in question. 

If you see errors pop up here, then it looks like you do actually have a disk issue. 

 

 

 

 

And specifically, I'm using one of these drives as a data drive on my desktop system (was the system disk), on an intel RST controller (Z97E chipset). 

[Prodigal]

v. 2.5.2.3121 BETA in the scanner settings window.

 

The burst test continues to run successfully without any errors for nearly 14 hours now. The flag that it gives only lasts for a very brief period. I happened to be sitting at my computer when I got another notification email (identical to the one quoted above), and pulled up the SMART data window on the drive immediately, it looked like this:

 

 

Less than a minute later, though, everything was back to being happy.

 

 

The drive in question is brand new SATA III SSD, connected to a Intel SATA III controller on an Intel H67 based motherboard. I do not currently have the Intel RST drivers installed on that machine. The drive is used as the OS/application drive only. It is not used for storage nor is it part of my Drivepool (which are all ~4-6 year old magnetic platter drives). I don't know if it's relevant at all, but before adding the SSD a couple weeks ago, the server OS was on a 60GB partition on one of the storage drives, which I cloned the boot & OS partitions to the SSD, then wiped the storage drive and returned it with a single partition back into the data pool.

[Christopher (Drashna)]

Actually, I just reproduced this issue myself.  The Current Pending Sector Count jumps up during a scan and then drops back to 0. 

 

It looks like this is a weird quirk with the firmware, or it's otherwise using the value wrong.  

And it happens very quickly. 

 

I've flagged this issue for Alex, so he can take a look at it. (eg, update the BitFlock rules for this device). 

 

https://stablebit.com/Admin/IssueAnalysis/27192

[Prodigal]

So nothing to worry about then?

[Christopher (Drashna)]

For now, yeah, nothing to worry about.  

 

It looks like heavy activity is triggering the issue (such as scans), and then it "clears up".  I've let Alex know about it, so we should be able to fix it on our end. But I can't really give an ETA.