What to do with HDD with bad sectors ?

[hhhd1]

I've got an 3 year old hard drive, 500gb 7200rpm, that I extracted from a laptop, as I was seeing its SMART details, I saw 4 re-allocated sectors.

Not sure what to do with this drive now, I am thinking of stress testing it to see if bad sectors increase, and if they do not, just use it as a backup drive or on home server for downloads.

What should I do with it ? How much can I trust it ?

How to stress test it ?

 

[Smoblikat]

You were right. Id run an HDTune on it, check SMART, even do a write test on it. If it looks stable enough, just use it as a storage/media drive. I wouldnt keep anything on it that cant be replaced though, treat it like the timebomb it is.

 

[dave_the_nerd]

Every hard drive is a time bomb.

A few reallocations are normal - you should monitor the drive regularly (because you check the SMART status every day... right? Right?!?) and if the number of reallocated sectors starts going up a lot, then at least you have a cool paperweight.

 

[pauldun170]

I had a HP Pavilion with a drive that reported errors after 2 years of use.
I continued to use that laptop with that drive for another 3 years.
Use it for stuff you back up regular

 

[flexy]

A low number of reallocated bad sectors are "normal". I have a bunch of backup drives which have 1 or 2 reallocated sectors and they work without issues for 2+ years.

SMART usually has a built-in threshold for nr. of reallocated sectors, once this number is exceeded SMART will alert you that failure is imminent and you need to backup your drive and replace as soon as you can. But no reason to worry if you only have 2-4 reallocated sectors and the number doesn't increase.

(Geek Tip: I have one very old drive which actually developed 1800 bad sectors due to unexpected power outages, obviously SMART kicked in telling me at boot to immediately replace the drive. I used a workaround trick to use the drive anyway, I re-partioned it and simply EXCLUDED and don't use the entire area at the back of the drive where the errors are. Of course that's VERY dirty workaround and I merely did it because I was curious and bored)

For testing.. check out HD Sentinel. It has options to do various tests on your HDD. Eg. it writes and re-reads all sectors and can tell you which sectors cause problems. The drive should actually automatically re-allocate bad sectors when it comes across them.

 

[hhhd1]

so, ..

I have done read / write / random access / benchmark / smart extended test, continuous for 10 hours.

Then I installed it in a laptop, installed windows 8 , ran windows update to finish up the drivers , copied over 80gb of data from USB.

At that point, the drive hangs, with the light continuously lit, tried rebooting into windows, it didn't work, just hang in the middle of loading.

Booted a linux USB, to find the SMART data saying reallocated sector count increased from 4 to 52.

Trying to perform full disk erase, via command
cat /dev/zero > /dev/sda
to start over. (just ran it for few minutes, then ctrl+c, only reseting the partition table)

restarted, this disk is seen in bios normally, but it wouldn't allow any windows , or even windows setup, to continue to boot at all.

restarted into linux boot usb, SMART data is now saying "Can not read SMART data".

So now, it is only visible in BIOS, and that is all.

 

[Elixer]

That isn't shocking....usually, at the first sign of SMART issues, it is time to replace the unit, and then use it as a scratch drive.

 

[CiPHER]

You guys look at Reallocated Sector Count. That is not important; those sectors are already swapped with reserve sectors.

The killer attribute to look for in the SMART data, is Current Pending Sector. That one is the real killer. It has to be zero (0) for raw value. Any non-zero raw value indicates an active bad sector that will make the application hang or produce errors. Any legacy filesystem could become corrupt with just one bad sector. Reallocated Sectors are harmless because they cannot be seen by the application or OS.

 

[tweakboy]

DisplayFusion . gl

HDD Generator is the dos app thru windows is good. It fixes bad sectors,,, I went from 16 ,, to 0 .......

 

[CoPhotoGuy]

DBAN and then recycle it.

 

[pitz]

I stick drives that might be going flaky like this in RAID-1 with a good drive. On Linux. If it fails to read a sector, then the system merely reads from the other drive(s) in the mirror without interrupting the system. That way, I'm still getting at least some use out of a drive which could very well carry on for many years of use.

 

[datarecoveryguy]

SMART is actually a terrible way to determine the health of a drive. By the time smart gets tripped and your computer says something like "smart check: fail." It's often too late and your drive is already in really bad shape. Your data may already be unrecoverable by simple means. While I may use HDD sentinel to do a quick smart check I prefer the manufacturers diagnostics. While they stress the drive more I find they are far more accurate. I've had drives that reported 100% in sentinel that fail completely days later. Back it up!

 

[CiPHER]

SMART is actually a terrible way to determine the health of a drive. By the time smart gets tripped and your computer says something like "smart check: fail."

SMART is useless this way indeed. You need to interpret the data yourself. If you can't, SMART means nothing to you. All applications that try to interpret SMART and give a health percentage, are simply bullshit.

But SMART is invaluable. You can instantly see the number of bad sectors and cabling errors, it is the first step of every diagnosis into your storage hardware.

 

[datarecoveryguy]

It was intended for the end user, but end users aren't going to monitor smart logs. Interestingly enough one single bad sector in the wrong spot can render the OS inoperable and even possibly make the drive useless. It would be much more useful if the obvious smart errors would show up to the end user weeks or months earlier. But that would cause a lot more warranty coverage for the manufacturers so we won't see that happening!

 

[BonzaiDuck]

It was intended for the end user, but end users aren't going to monitor smart logs. Interestingly enough one single bad sector in the wrong spot can render the OS inoperable and even possibly make the drive useless. It would be much more useful if the obvious smart errors would show up to the end user weeks or months earlier. But that would cause a lot more warranty coverage for the manufacturers so we won't see that happening!

All the advice so far is useful. But hard drives today are so cheap, and enthusiast collections of older drives so likely to be plentiful, I'd do this.

Get your Weatherby .457 Magnum bolt-action elephant gun out of the closet -- clean it up a bit; take it with the HDD to the gun range; set up the HDD a couple hundred yards away so it reflects bright sunlight; stick one of those hand-grenade-sized rounds in the chamber. Then fire away!!

If you leave it at the gun range after scoring a hit, nobody should be able to extract any data from it.

But if you didn't go equipped with a shoulder pad, drop by the drug-store on the way home to stock up on Ben-Gay and pain-pills.