Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

This question already has an answer here:

I have a seagate 320 GB hdd on my linux machine. From a week, SMART utility has been warning me to backup as the disk failure is imminent.

Report failed on the "Bad sectors attribute" which is Normalized: 1 Worst: 1 Threshold: 36 Value: 4095 sectors

I don't know the meaning of these terms. Can someone please explain and suggest me what to do? Will reformatting the entire hard disk help?

P.S. I can't post the images of the SMART report since my reputation points are less than 10.

EDIT: How can I tell when will the disk fail? Any measures to extend the life by some weeks?

share|improve this question

migrated from serverfault.com Sep 7 '13 at 18:37

This question came from our site for professional system and network administrators.

marked as duplicate by techie007, Mokubai, Dave M, Breakthrough, Scott Sep 9 '13 at 23:08

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

10  
1) Back up your stuff (quickly), 2) Buy a new drive, 3) Install new drive, 4) Restore your stuff –  squillman Sep 7 '13 at 14:06
1  
You cant tell WHEN it will fail it already HAS FAILED. There is no data that can be generated to determined when exactly it wont work at all –  Ramhound Sep 7 '13 at 19:01
    
Your SMART is set to warn you of potential failure when your Bad Sector count hits 36 (because it's out of replacement sectors to use), and you're are at 4095. So, as the drive is telling you, it's failing and can't "fix" itself any more, replace it ASAP or expect data loss. –  techie007 Sep 7 '13 at 22:12
    
@techie007 If the disk has run out of spare sectors, technically there already is some degree of data loss. Hopefully it is in unused parts of the disk, but if that was the case, (absent recent full disk SMART tests) why would the disk be warning about them? –  Michael Kjörling Sep 9 '13 at 11:36
add comment

5 Answers

Hard disks work by writing data to spinning magnetic disks. The disks are divided into tracks (concentric rings). The tracks are divided into sectors (sections around the ring).

Your utility is telling you that several sectors have gone bad (it can't properly read/write data on those sectors).

Now HDDs generally have a certain ability to compensate for bad sectors. It may have error correction, it may have spare sectors, etc, etc. Your HDD has run out of tricks, and so you are seeing the bad sectors at a higher level.

GET RID OF THE DRIVE. Yes, it IS going bad, no you CAN'T fix it.

share|improve this answer
    
Why did they disk fail so badly? I mean I have read a lot online about this problem and the maximum bad sector count I came across that a person got was about 400-500. Is this caused by a malware? –  gargie Sep 7 '13 at 16:25
    
@Gargee Disks go bad; it happens. It's not malware. Malware doesn't cause hardware failures. Defects and other physical actions (power surges, blunt force, etc) cause hardware failures. –  squillman Sep 7 '13 at 16:48
    
@squillman Ok. And how can I tell when will the disk fail? Any measures to extend the life by some weeks? –  gargie Sep 7 '13 at 17:03
3  
@Gargee This is silly. In a professional environment a SATA hard drive is so cheap that it is essentially free. Why didn't you replace it yesterday? There is nothing to be gained by extending the life of such an inexpensive device. –  Miles Erickson Sep 7 '13 at 17:08
4  
@Gargee - the unusual thing here isn't that you have so many. It's that the drive is still useable at all at this point. It won't stay that way. THERE IS NO HOPE FOR THIS DRIVE. Your only hope now is to get your data out before it's too late. –  Michael Kohne Sep 7 '13 at 17:08
show 1 more comment

In my experience with heavy use design workstation, I have learned the "burning way" to definitely save a lot of money spending few of it.

If your S.M.A.R.T. test of the Hard Disk says that a failure is imminent, it means so.

As a matter of experiment, I tested using UBUNTU Distro three HP-540D Workstations, whose hard disk have been almost continuosly spinning for more than six years.

I had them on a shelf, to be used as emergency "backup workstation", but the reliability of the "backup" is also a very important matter.

So i decided to transform the " Backup " in "Hot Standby" using a Pen Booting Distro, and keeping an eye on the S.M.A.R.T. Status.

Within months from startup, all three workstations behaved in the same way: the status turned to " a failure is higly probable in the next 24 hours.

All three hard disks failed , in some way, within two months from the first warning.

This does not mean that the Status warns you WITHIN two monts from the event. This means that at least three HD kept running for less than two months from the first warning.

As a System Administrator is a totally different music.

My advice is , as a strong and urgent suggestion, to replace at the EARLIEST convenience the failing device.

As US Marines usually say: " You have been WARNED!", and it's last time they say somethingh before they shoot you. Nobody can say how long it takes to shoot you, but "You have been WARNED!"

Enjoy your week end.

EDIT: The HD is doomed! You MAY be so lucky to have the time to "image" it and then replace with no data loss. Please , understand that HD are non mass commodities, with a MRSP less than 100 USD.

share|improve this answer
add comment

Although most of the generic stuff has already be said, but formally the specific answers to your questions are missing:

I don't know the meaning of these terms. Can someone please explain and suggest me what to do?

Your device has irrecoverably lost a smaller amount of data. Statistically, it is likely that more is to follow. You also are at a greater risk of losing the entire drive ("greater" as in "a higher probability than a comparable drive population without bad sectors would expose").

Will reformatting the entire hard disk help?

No.

How can I tell when will the disk fail?

You can't. SMART is guessing, it can't be sure of anything. There has been a research at Google concerning HDD failures with a rather fuzzy result about SMART predictions:

Our analysis identifies several parameters from the drive’s self monitoring facility (SMART) that correlate highly with failures. Despite this high correlation, we conclude that models based on SMART parameters alone are unlikely to be useful for predicting individual drive failures.

So basically, a drive with SMART errors might run for another year or might fail completely the next minute. Statistically, it is more likely to fail than a drive without SMART errors, but this will say nothing about how your drive will behave.

Any measures to extend the life by some weeks?

None except from switching it off and not using it for the said number of weeks. If it's broken, it might break even more and take the data with it.

share|improve this answer
1  
As for the last point: depending on the type of failure, it may actually make the situation worse if you turn the drive off compared to keeping it spinning 24/7 until you can get the data off the drive. Consider for example the possibility of the spindle motor for some reason having trouble keeping a stable speed, and that this causes difficulties for the drive to read and write data. In such a case, it might not come back to life if shut down. –  Michael Kjörling Sep 10 '13 at 11:30
    
@MichaelKjörling you are right, although bad sectors due to surface defects seem to be more common. –  syneticon-dj Sep 10 '13 at 13:37
    
I agree, it's a slightly contrived situation. My point was simply to illustrate that shutting the drive down might not always be the best thing to do when it is failing, depending on the failure mode. Hence may and for example. –  Michael Kjörling Sep 10 '13 at 15:08
add comment

SMART is letting you know that you've encountered a physical disk error. Although you could place a new filesystem on the drive this will not help the physical problem that you're having. You should get a good backup and replace the faulty drive so that your data remains safe.

share|improve this answer
add comment

I want to elaborate on your question of whether or not you can extend the life of your hard disk. I too have a 320GB Seagate drive that has been giving out on me, but I'm still using it a year later. I understand that the proper way to deal with such errors is to replace the drive, but not everyone has the resources to do so.

Here's what you can do:

  • Remove all crucial data to external drives if you have the option to

  • If you sport a journaling file system then keep the file count on your drive as low as possible. Delete all files you already have in archives. Uninstall unused programs.

  • Disable preloading mechanisms if they exist.

  • If you have enough RAM then disable the swap file

  • As people have pointed out, your hard drive has unrecoverable sectors. If you stumble over the file(s) that occupy them, then don't delete them! This way these sectors will stay in use. The files will be useless but with some luck new ones won't pop up that quickly and functioning files won't end up being accidentally written to them.

  • You could further lower the strain on the drive by using a live disc or boot the OS from a stick.

On my drive these tactics have worked out fine. I merely store absolutely replaceable data on it until it either crashes beyond any point of recovery or I can afford a replacement drive. In any case, SMART is not reliable. I had plenty of hard drives die without it ever having been triggered. On the other hand, a SMART message doesn't necessarily mean your hard drive will die tomorrow, but the chances are certainly higher than for drives that check out okay.

share|improve this answer
add comment

Not the answer you're looking for? Browse other questions tagged or ask your own question.