Hard drive is acting up but passes SMART/WD Diagnostics with no errors...

[lektrix]

What I mean by my drive "acting up" is that everytime I reboot, it asks me do a chkdsk for this specific drive...there are always "corrupt" files that need to be deleted or realigned or w/e and it happens on EVERY bootup. Recently, I've had some of my game folders disappear but when I click the shortcuts on my desktop, the games still work fine...

It seems to me like the drive is dying (Raptor 74GB - out of warranty), but it passes the SMART test and has 0 errors in a WD Diagnostics Extended Test. Error scan in HD Tune 4.01 shows nothing as well, but when I got to 'Health', it says:

(05) Reallocated Sector Count 199(Current)/199(Worst)/140(Threshold)/3(Data) - warning
(C4) Reallocated Event Count 199/199/0/1 - warning
(C7) Ultra DMA CRC Error Count 200/253/0/308 - warning

What does this mean? Is the drive dying or what?!@?!

 

[MagnusTheBrewer]

What OS are you running? Try running sfc /scannow

 

[bkch0]

the drive is developing bad blocks.

when bad blocks are found, they are no longer used and this logical space is replaced with reserved blocks.

When this happens, it shows on smart rellocated sector counts

 

[SparkyJJO]

FWIW, WD Diag is totally worthless IMO. I've had obvious failing drives (horrible transfer rates, click of death, etc) and it passes them fine.

SMART isn't too bright either.

 

[n7]

1.
Speedfan > SMART tab > select HDD > Perform an indepth online analysis.

2.
Post link here.

3.
Profit (or in this case, likely, fail )

 

[sub.mesa]

Do you have any pending sectors? Reallocated sectors means the HDD fixed its own damage and the bad sectors are not in use anymore and cannot cause damage of any kind.

However, any pending sector means that the HDD has a bad sector that has ACTIVE data on it; that means it will cause troubles when used in RAID and Windows will most likely panic and put the drive in PIO very quickly. Always look for pending sectors; they are dangerous and even 3000 bad sectors is harmless compared to 1 pending sector.

 

[Emulex]

malware? are you shutting down cleanly?

 

[Russwinters]

05) Reallocated Sector Count 199(Current)/199(Worst)/140(Threshold)/3(Data) - warning

This field only DATA is relevant. The drive has located 3 bad sectors.

(C4) Reallocated Event Count 199/199/0/1 - warning

DATA is the only relevant field again, there was 1 reallocation event

(C7) Ultra DMA CRC Error Count 200/253/0/308 - warning

CRC errors.....check your cable and you storage controller (motherboard) for instability. There could be something going on with the communication of the drive -> storage controller.


Run this:http://hddscan.com/


The "MAP" view helps get a visual idea of where the problem areas on your drive are.

 

[lektrix]

I apologize for opening two threads (accident).

I plugged the Raptor into another working port and working cable and it "works" now. Mind you it still does the customary chkdsk before I enter Windows, but HD Tune benchmarks were finally normal.

SpeedFan Online Analysis - http://www.hddstatus.com/hdrepshowreport.php?ReportCode=4380159&ReportVerification=529DA954

And screenshot of HDDScan and HDTune: http://img195.imageshack.us/img195/8661/wd74gbraptorfla1.jpg

I can access the drive but there are missing files/folders.......they aren't important and I got what I need, so my final question is this:

Is it still safe to use this drive?

 

[RebateMonger]

Is it still safe to use this drive?

NO drive is "safe to use". They all fail eventually, sometimes with no warning. That's what backups are for.

If you end up convinced that cabling was the only problem, then, sure, use the disk. But if it looks like an internal disk error, I'd replace it. My experience with "fixing" disks hasn't been good. More errors will occur. Maybe an hour later, maybe a day later, maybe a month later.

 

[lektrix]

Yes, wiped it in Windows and wrote zeroes to it (full format). Now the drive is "dying". I can barely run any benchmarks, read speed would drop to 2.0MB/s and Error Scan shows 62.2% damaged blocks! WTF.

(05) Reallocated Sector Count 3
(c4) Reallocated Event Count 1
(C7) Ultra DMA CRC Error Coutn 352

Did I do the reformat wrong?

(I also changed to AHCI mode, guys)

 

[n7]

Formatting doesn't magically fix a dying HDD.
It can delay things a bit longer allowing for a bit longer usage, but that's all.

Based on your HDDScan results (see the 47 "yellow" results), it's effed.
There's no way a Raptor should be getting anything in the yellow area for access time.

You've also got at least a few reallocated sectors, with many more on the way.

Your drive = dying.
Get your stuff off it, it's smashing time.

 

[Modelworks]

HDDSCan will show you the hex values for the SMART values.
Using windows calculator in programmer mode put in the hex values and then select DEC to see the real SMART value .

For example my drives report says:
PowerOn Hours Count 0000 for the SMART value, which makes no sense, but the hex for that function says :
PowerOn Hours Count 2B6A in hex
Converted that is 11,114 which makes sense , the drive is about 3 years old so that is 412 days total .

The next thing you can do is surface test in the task menu. Click verify in the menu and add task and let it run.

 

[lektrix]

HDDSCan will show you the hex values for the SMART values.
Using windows calculator in programmer mode put in the hex values and then select DEC to see the real SMART value .

For example my drives report says:
PowerOn Hours Count 0000 for the SMART value, which makes no sense, but the hex for that function says :
PowerOn Hours Count 2B6A in hex
Converted that is 11,114 which makes sense , the drive is about 3 years old so that is 412 days total .

The next thing you can do is surface test in the task menu. Click verify in the menu and add task and let it run.

On the right image HDTune says 31590 hours which is roughly 3.6 years. The warranty for this drive just finished at the end of 2009...so not bad....but I've seen hard drives last 7-10 years without a hitch though and this is a freakin' RAPTOR!

I can't believe I had two drives die within 2 weeks...

Does this drive failure have anything to do with me hot swapping another drive (which is now dead and clicking) without hot swap enabled on the machine? Basically I removed another drive and reinserted another drive while Windows was running 2 weeks ago (it has nothing to do with this drive though)...

 

[lektrix]

Formatting doesn't magically fix a dying HDD.
It can delay things a bit longer allowing for a bit longer usage, but that's all.

Based on your HDDScan results (see the 47 "yellow" results), it's effed.
There's no way a Raptor should be getting anything in the yellow area for access time.

You've also got at least a few reallocated sectors, with many more on the way.

Your drive = dying.
Get your stuff off it, it's smashing time.

What's an acceptable number of "yellow" results? I'm running the scan on all 5 of my HDDs to see which ones need to be replaced..

 

[Russwinters]

Yah, I wouldn't use this one anymore. You could probably pull alittle bit more time out of it by using some programs that for sector reallocation, but it's just not really worth it.


A hard drive lasting 7-10 years is good luck.


I am happy if my drive works without a hitch for 3-4 year, at that point i replace it (and I always keep a backup of important things)


Regards,

 

[lektrix]

I'm running the HDDScan on two other drives and so far:

WD74GB OS Drive (my second Raptor, no warranty) has 6 yellows and it's been powered on for 3.5yrs

WD500GB AAKS Data Drive (1yr warranty left) has at least 10 yellows (still scanning) and it's only been powered on for 1/2 year

 

[n7]

While the HDD Scan program in Windows is decent, i really wouldn't want to rely on it too much as while in the OS, many things can affect the test.

This the bootable version of it: http://hddguru.com/content/en/software/2005.10.02-MHDD/

MHDD generally will only see one HDD @ a time, the "Master" HDD only.

So for SATA, you need to be in IDE mode, & might need to be using SATA port 1 only, but the program is great for checking HDDs.

Once you've booted into MHDD & selected the HDD, F8 shows SMART data, & F4 then F4 again will run the scan.
Same idea as the HDD Scan one in Windows, but being bootable means other things aren't possibly interfering/causing poorer access times.

Over 150ms time isn't really good; ideally you want lower access time basically.
The box on the bottom right corner (of the MHDD scan windows) will show information about bad stuff if it finds it.

 

[lektrix]

A little bit off-topic, but is this drive about to die soon as well?

http://img69.imageshack.us/img69/8751/seagatem.jpg

 

[Russwinters]

MHDD is an amazing program; you will notice it is written by Dmitry Postrigan, who is actually my Friend/Guru/Mentor/Trainer.

That seagate drive has a lot of CRC errors...could be a cable/controller problem.


3 reallocated isn't anything serious YET, but it is what SMART is designed for...to raise a flag if abnormal there are abnormal occurrences.

a few yellows is OK, that just means an area of the disk is degrading...if you have a few yellows (under 20 or so) i would not worry (you should ALWAYS have a backup anyways)

What is a serious indicator of a problem is a pattern like this:

http://www.atola.com/img/insight_pics/media_scan_bad_head.png

This SS is taken from Atola insight, which is the evolution of MHDD into a professional Data recovery tool. This is what I use in my office. You can use MHDD to see these patterns as well, looking for yellow/orange/red blocks in a pattern (like every few thousand sectors you hit another group of sectors)

This indicates that one or more heads are going bad, but the H0 (usually the bottom head) is probably ok because the drive could ID and read the firmware.