S.M.A.R.T. HDD failure: how to repair?

[axelfox]

UPDATE 4: Ran WD's write zero program. It fails 3/4 into it. I think this sucker is done.

UPDATE 3: The drive is probably set up to be a goner. I backed everything up and trying to ovewrite with zeroes as a last resort. The warranty expired 12/14/04. Thanks everyone.

UPDATE 2: I ran WD diagnostics full test and it supposedly fixed the bad sectors.

Upon reboot, however, the drives aren't detected. I remove the cables and then put them back in and reboot.

I get the SMART failure again and run the WD diagnostics, only to tell me that the SMART has failed again.

I ran EVEREST with the results in the post below.

At this point I have backed up my data on another drive. I can low level format if that will fix the bad sectors (or what ever is wrong).

_________

My primary slave started to get this message upon detection of the IDE channels:

Primary Slave: S.M.A.R.T. hard drive failure. Failure imminent. Back up and replace.

Its been doing that for a couple of weeks and I'm waiting for it to crash. Does anyone know how to fix it?

Thanks

UPDATE BELOW.

 

[harpy82]

"... for a couple of weeks and I'm waiting for it to crash. Does anyone kno..."

so... the hdd has not yet crash ?

S.M.A.R.T detection isn't really tat all smart... so.. i guess is best is to disable S.M.A.R.T to get rid of the error messages.... then get a new drive and transfer ur data over.. drives r so cheap this day that i wouldn't worri about upgrading one... unless it has already failed... then i will sweat... cry...... suicide... LOL

 

[montag451]

I think the way SMART works is on probabilities. If a certain event happens, then you get the warning if the probability of a crash is imminent to not so imminent.
There is no way it can determing when your drive is going to give up for definite.
You do need to backup/clone for when the big bang does happen.

 

[WobbleWobble]

I've found that zero-filling ("low level formatting") the drive gets rid of some SMART errors. That will kill your data on the drive though.

 

[montag451]

Maybe it would be worth doing this is a last resort - Ghost the drive/bit by bit. then zero-fill the old drive.
At least you should be able to recover the data one way or another -
SPINRITE 6, GetData Back, Recover4All, Recovery2000

 

[Jeff7]

Download and install Everest. Run it, go to its Storage section, and click on SMART. Right click on the attributes of the drive that's causing the problems (hopefully it's visible there), click Copy All, and paste the results back here.

 

[CrispyFried]

SMART keeps track of bad sectors and read/write retries, no? It may be his bad sector relocation table is almost full. I would back up and replace, why take chances.

 

[FlyingPenguin]

SMART keeps track of a couple of dozen HDD parameters. If any of these parameters go out of factory spec, SMART gives a warning. It's meant to give you some advanced warning of an impending drive failure.

It's stupid to disable SMART. It does no harm, and can warn you of a drive failure. Turning off SMART is like removing the "Engine Service Soon" light from your car's dash.

SMART tends to err on the safe side. You can get SMART errors and still have a perfectly good drive - it's just SLIGHTLY out of factory spec.

Sometimes a drive will give a SMART error when you boot and sometimes it won't - it's only going out of spec once in a while.

This is NOT something you want to ignore. Your HDD might be failing and you could lose your data.

Download the HDD manufacturer's diagnostic utility from their website and run a full (Advanced) diagnosic on the drive. If the drive fails the diagnostic you should backup your data immediately and replace the drive.

 

[Jeff7]

Originally posted by: FlyingPenguin
SMART keeps track of a couple of dozen HDD parameters. If any of these parameters go out of factory spec, SMART gives a warning. It's meant to give you some advanced warning of an impending drive failure.

It's stupid to disable SMART. It does no harm, and can warn you of a drive failure. Turning off SMART is like removing the "Engine Service Soon" light from your car's dash.

SMART tends to err on the safe side. You can get SMART errors and still have a perfectly good drive - it's just SLIGHTLY out of factory spec.

Sometimes a drive will give a SMART error when you boot and sometimes it won't - it's only going out of spec once in a while.

This is NOT something you want to ignore. Your HDD might be failing and you could lose your data.

Download the HDD manufacturer's diagnostic utility from their website and run a full (Advanced) diagnosic on the drive. If the drive fails the diagnostic you should backup your data immediately and replace the drive.

Except that SMART is more advanced than a Check Engine Light.
A bit of a rant here - I don't like the Check Engine Light. It's like having a light on the front of a PC that reads "General Computer Failure" if anything goes wrong. Sure narrows it down, doesn't it? :roll:
Be nice if all cars had a little LCD built in that would display the error code instead of just lighting up a single light. Epox built POST cards into their motherboards - cars should come with the same thing.

Originally posted by: CrispyFried
SMART keeps track of bad sectors and read/write retries, no? It may be his bad sector relocation table is almost full. I would back up and replace, why take chances.

It keeps track of that, as well as lots of other things.
Here's a sample of what I get:

ID Attribute Description Threshold Value Worst Data Status
01 Raw Read Error Rate 6 60 54 20936006 OK: Value is normal
03 Spin Up Time 0 96 96 0 OK: Always passing
04 Start/Stop Count 20 100 100 94 OK: Value is normal
05 Reallocated Sector Count 36 100 100 0 OK: Value is normal
07 Seek Error Rate 30 80 60 106431477 OK: Value is normal
09 Power-On Time Count 0 98 98 2241 OK: Always passing
0A Spin Retry Count 97 100 100 0 OK: Value is normal
0C Power Cycle Count 20 100 100 101 OK: Value is normal
C2 Temperature 0 36 46 36 OK: Always passing
C3 Hardware ECC Recovered 0 60 54 20936006 OK: Always passing
C5 Current Pending Sector Count 0 100 100 0 OK: Always passing
C6 Off-Line Uncorrectable Sector Count 0 100 100 0 OK: Always passing
C7 Ultra ATA CRC Error Rate 0 200 200 0 OK: Always passing
C8 Write Error Rate 0 100 253 0 OK: Always passing
CA <vendor-specific> 0 100 253 0 OK: Always passing


Not exactly well formatted, but the data can be gleaned from that. This is a Seagate drive. Note all the read and seek errors, but also look at the Hardware ECC Recovered number. It's always the same as the Read Errors. Modern hard drives seem to screw up all the time, but they are able to figure it out on the fly, and produce good data instead of a stream of corrupted bits.

 

[imported_Phil]

Originally posted by: FlyingPenguin
SMART keeps track of a couple of dozen HDD parameters. If any of these parameters go out of factory spec, SMART gives a warning. It's meant to give you some advanced warning of an impending drive failure.

It's stupid to disable SMART. It does no harm, and can warn you of a drive failure. Turning off SMART is like removing the "Engine Service Soon" light from your car's dash.

SMART tends to err on the safe side. You can get SMART errors and still have a perfectly good drive - it's just SLIGHTLY out of factory spec.

Sometimes a drive will give a SMART error when you boot and sometimes it won't - it's only going out of spec once in a while.

This is NOT something you want to ignore. Your HDD might be failing and you could lose your data.

Download the HDD manufacturer's diagnostic utility from their website and run a full (Advanced) diagnosic on the drive. If the drive fails the diagnostic you should backup your data immediately and replace the drive.

Top advice :thumbsup:

 

[axelfox]

Ran WD diagnostic, and might be able to repair. However, I'll have to back up 60 gigs worth of data on DVDs before I can repair it so it doesn't delete the data.

 

[imported_Phil]

Originally posted by: axelfox
Ran WD diagnostic, and might be able to repair. However, I'll have to back up 60 gigs worth of data on DVDs before I can repair it so it doesn't delete the data.

I wouldn't trust that drive again, personally, but......

 

[axelfox]

UPDATE ABOVE

 

[axelfox]

Originally posted by: Jeff7
Download and install Everest. Run it, go to its Storage section, and click on SMART. Right click on the attributes of the drive that's causing the problems (hopefully it's visible there), click Copy All, and paste the results back here.

Here are the results: BOLD ADDED

ID Attribute Description Threshold Value Worst Data Status
01 Raw Read Error Rate 51 200 178 18 OK: Value is normal
03 Spin Up Time 21 89 78 6308 OK: Value is normal
04 Start/Stop Count 40 99 99 1634 OK: Value is normal
05 Reallocated Sector Count 140 100 100 1590 Pre-Failure: Imminent loss of data is being predicted
07 Seek Error Rate 51 200 200 0 OK: Value is normal
09 Power-On Time Count 0 72 72 20585 OK: Always passing
0A Spin Retry Count 51 100 100 0 OK: Value is normal
0B Calibration Retry Count 51 100 100 0 OK: Value is normal
0C Power Cycle Count 0 100 100 495 OK: Always passing
C4 Reallocation Event Count 0 1 1 818 OK: Always passing
C5 Current Pending Sector Count 0 200 200 2 OK: Always passing
C6 Off-Line Uncorrectable Sector Count 0 200 200 4 OK: Always passing
C7 Ultra ATA CRC Error Rate 0 200 253 4 OK: Always passing
C8 Write Error Rate 51 200 200 3 OK: Value is normal

 

[FlyingPenguin]

Once you have bad sectors on a modern HDD, death is imminent. RARELY does zeroing out the drive fix this anymore. That's old skool fix. Modern drives have such a small head gap that ANY debris from the initial head crash will quickly lead to a domino effect of more head crashes that in turn add MORE debris, and so on.

Use the drive as little as possible until you back it up. I would strongly suggest you buy another drive and just copy it all over. I wouldn't use that old drive for anything mission critical.

You should check the manufacturer's online warranty checker to see if it's still covered and if so, you can RMA it.

 

[axelfox]

Originally posted by: FlyingPenguin
Once you have bad sectors on a modern HDD, death is imminent. RARELY does zeroing out the drive fix this anymore. That's old skool fix. Modern drives have such a small head gap that ANY debris from the initial head crash will quickly lead to a domino effect of more head crashes that in turn add MORE debris, and so on.

Use the drive as little as possible until you back it up. I would strongly suggest you buy another drive and just copy it all over. I wouldn't use that old drive for anything mission critical.

You should check the manufacturer's online warranty checker to see if it's still covered and if so, you can RMA it.

So I guess I don't have anything to lose by zeroing out the drive, huh? I've already backed everything up.

 

[axelfox]

UPDATE 3.

 

[FlyingPenguin]

So I guess I don't have anything to lose by zeroing out the drive, huh? I've already backed everything up.

Nothing to lose, but I personally wouldn't trust that drive with anything on it you can't afford to lose in the future.

 

[pkypkypky]

some friendly advice...

Backup important data and start looking for the right HDD deal for yourself. Then put up your HDD on ebay and note that the HDD is having Smart detection errors and you will not be responsible for failure.

And most importantly, make sure your next drive has a 3-5 yr warranty. Also, save your receipt to make sure you get the full warranty period. Good luck.

 

[axelfox]

I jumped in on the Seagate 300GB for $120 AR. Come with 5 year warranty.

Thanks for the help!

 

[Jeff7]

Originally posted by: axelfox

Originally posted by: Jeff7
Download and install Everest. Run it, go to its Storage section, and click on SMART. Right click on the attributes of the drive that's causing the problems (hopefully it's visible there), click Copy All, and paste the results back here.

Here are the results: BOLD ADDED

ID Attribute Description Threshold Value Worst Data Status
01 Raw Read Error Rate 51 200 178 18 OK: Value is normal
03 Spin Up Time 21 89 78 6308 OK: Value is normal
04 Start/Stop Count 40 99 99 1634 OK: Value is normal
05 Reallocated Sector Count 140 100 100 1590 Pre-Failure: Imminent loss of data is being predicted
07 Seek Error Rate 51 200 200 0 OK: Value is normal
09 Power-On Time Count 0 72 72 20585 OK: Always passing
0A Spin Retry Count 51 100 100 0 OK: Value is normal
0B Calibration Retry Count 51 100 100 0 OK: Value is normal
0C Power Cycle Count 0 100 100 495 OK: Always passing
C4 Reallocation Event Count 0 1 1 818 OK: Always passing
C5 Current Pending Sector Count 0 200 200 2 OK: Always passing
C6 Off-Line Uncorrectable Sector Count 0 200 200 4 OK: Always passing
C7 Ultra ATA CRC Error Rate 0 200 253 4 OK: Always passing
C8 Write Error Rate 51 200 200 3 OK: Value is normal

Lessee....Got a good number of Raw Read Errors, Reallocated sectors, as you can tell, are high, Reallocation events are also quite high, a few Pending Sectors, Uncorrectable Sectors, ATA CRC Errors, and some Write Errors. Yes, there is definitely much about this drive that is seriously borked.

Fairly old drive too, it seems. 1634 power cycles, power-on time is at 20,585 (hours I believe that is).


Thing that stinks about SMART - it'll start showing reallocated sectors, or perhaps some other errors, but if it's within tolerance, it won't say a thing. I get nervous if it shows any reallocated sectors, especially if that number starts to increase gradually. My dad's one hard drive is showing 8 reallocated sectors, but the drive doesn't think it's anything serious. I'll keep an eye on that number though.