Bizarre S.M.A.R.T Behavior on Disk..including "Worst" going UP

Rick441

Distinguished
Apr 10, 2011
24
0
18,520
Four days ago I got a Windows 7 Popup about how S.M.A.R.T reports that hard disk failure might be imminent for my laptop’s 4-year-old HGST 1TB drive. Neither Speedfan’s SMART report nor Western Digital’s Diagnostic (which doesn’t even give raw values) flagged any issues, but HDDScan had a caution symbol (yellow triangle with exclamation point) next to Reallocation Event Count, even though both the Current and Worst SMART values still read 100. (BTW, there is no threshold value given for Reallocation Event Count.)

But I looked at the raw value, which was in Hex format, and using a Hex-to-Decimal converter saw that it was 29. Based on some things I read about even 30 perhaps being an iffy raw value for this measure, I placed an online order for a replacement drive. Over the next few days while awaiting the new drive, I kept tabs on this raw value, and it kept rising fairly steadily, to 54. This seemed to confirm that there could be trouble brewing with the drive (although the only *possible* symptom I could detect was occasional brief hesitation before responding to typing or scrolling). (BTW, even at that point Speedfan SMART and WD Diagnostics SMART still had the drive passing with flying colors. Oddly, there is no diagnostic from HGST that recognizes their own drive!)

But then things got weird: Last night, the night before the new drive was to arrive, I got two more Windows SMART disk warnings in quick succession. And an HDDScan SMART listing run immediately afterward showed the top line, Raw Read Error Rate, now had Current and Worst SMART values of 60 (down from 100), on a threshold of 62. So there was a “Critical” symbol next to it (X in a red circle). (Reallocation Event Count was still raw 54.)

I immediately did another backup (imaged the drive), but when I was done, both the Current *and Worst* SMART values for Raw Read Error Rate had jumped *up* to 75 on HDDScan, and there was no longer even a Caution symbol for the value, let alone Critical. I kept using the computer, and an hour later, both the Current *and Worst* values had jumped back up to 100! An hour later, they had both fallen to 85.

I have the new drive in now, but any thoughts on why the heck I could have been getting those weird jumping-around SMART readings for Raw Read Error Rate -- which Windows was apparently taking seriously? In particular how can WORST values go UP??? In this case WAY up. I’ve been Googling, and find no reports of that kind of behavior.

Thanks.
 
Solution
i going to be a little tech here. used to work for emc storgae before health got me. on new hard drives there two list that you cant see a plist and a glist. the plist is used to mask out known bad spots on any hard drive so there not used. the glist is for when a spot goes bad after the drive leaves the factory. the glist will flag smart after a set size of bad spots are found. with glist you can do a low level format and with a few free tools reset it if the groth stops and the drive itself is fine. with your drive the glist was growing and the drive was repairing or masking some of the head crashes/spindle stalls. when you felt the drive stall and stop it was the start of the dreaded wd spindle crash. when a wd dies it spindle locks...

atomicWAR

Glorious
Ambassador
Failing drives numbers for SMART can fluctuate wildly when a drive is going bad but some flux can be normal too in the middle of a drives life. You made the right call move getting it replaced when you did as your numbers were all over the place not just a few points up or down. Generally speaking keeping an eye on your hex data in SMART is always a good idea so you can spot these fluctuations early or conversely see and see if they are serious or just a drives mid life crisis so to speak.
 
i going to be a little tech here. used to work for emc storgae before health got me. on new hard drives there two list that you cant see a plist and a glist. the plist is used to mask out known bad spots on any hard drive so there not used. the glist is for when a spot goes bad after the drive leaves the factory. the glist will flag smart after a set size of bad spots are found. with glist you can do a low level format and with a few free tools reset it if the groth stops and the drive itself is fine. with your drive the glist was growing and the drive was repairing or masking some of the head crashes/spindle stalls. when you felt the drive stall and stop it was the start of the dreaded wd spindle crash. when a wd dies it spindle locks and the drive starts beeping. at that time your data is gone.
 
Solution

atomicWAR

Glorious
Ambassador
^much more technical but what I was getting at with some flux being normal but near the end the numbers get really jitttery up and down. As he stated the drive locks up when it dies. Great job keeping it understandable Smorizio but adding a lot of detail I breezed past.
 

Rick441

Distinguished
Apr 10, 2011
24
0
18,520
Great info smorizio and atomicWAR...really helps explain what could have been happening. Thanks. I'll wait a few days until marking one the solution, in case anyone else wants to add personal experiences or other ideas.

 

Rick441

Distinguished
Apr 10, 2011
24
0
18,520
Thanks again to both for the info which helps a lot in explaining what I was seeing...plus some additional interesting detail about what's actually going on "behind the scenes." One thing I do wonder about: Namely, that with my replacement drive I have still encountered a few instances where -- when typing in Firefox -- there is a brief hesitation before responding to typing. I.e., the cursor won't even "anchor" in the textbox for a few seconds. So maybe that was related to the browser or the specific site, rather than the hdd? The same thing had been happening to me in Word once in awhile with the old drive, but not with the new one -- at least so far.