Bad sectors on HDD help

I have had my inspiron 1525 for a little over a year. The HDD is a samsung 160 gb. Anyway, i booted into the new ubuntu 9.10 that i installed on a flash drive, and a pop up came up stating that it had many bad sectors and that i should replace my HDD soon. I clicked for more info, and found that it says i ahve over 4000+ bad sectors as of this writing. However, in windows, when running disk error check and other utilities, no such problem exists. How can I check for sure if I have a bad HDD, and if either windows or ubuntu is incorrect?

 

Does the hdd smart info say you have bad sectors, have you tried the free hdtune to check smart data and do bad sector scan.

 

You can run chkdsk (checkdisc) - and use crystaldiskinfo
http://crystalmark.info/software/CrystalDiskInfo/index-e.html to read SMART data.

If you have carried your laptop around a lot though, I could imagine there being a few bad sectors.

 

Is Ubuntu complaining about the hard drive or the flash drive?

Go grab stressLinux or something and let it have a go at your hard drive.

Don't assume that you know better than any non-windows diag program. If you have a diag program that says your hard drive is on it's way out, Read and Believe.

Do NOT try multiple diag programs to try and find one that says your hard drive is ok.

 

This is what crystal HD info shows. i guess i'm in trouble.

i was looking into a new HDD and found a seagate 160 GB 7200 rpm. guess that'll be a good replacement.

 

Would be ideal to replace it.

My grandfather had one of our old laptops with a HDD that read "caution" - and it worked well for over a year - but this means your HDD is no longer in "top form" - it would be better to replace it.

(Unless you want to risk data loss, then you can continue to use it)

 

Before you continue stress testing the HD, it might be wiser to back up your personal data first.

 

everything's backed up! thanks for the advice everyone. i'll ride it out until it dies, then order another one. luckily, i have no more english classes, so i won't need my laptop at school other than surfing the internet.

 

Update.

I ran that crystal disk again. now have 3 more bad sectors. that is in a time span of only a few hours.... is this normal?

and how many sectors does a 160 GB HDD have?

 

The more you use it, the more it will be damaged. So yes, it is normal. I'd replace the drive now.

 

327684 give or take.

 

Nah. Bad sectors aren't a worry. It's when you get a seek error or a sector reed error that you should be worried.

 

Backup all the data, wipe the drive with DBan or similar wiping program, and see if it fixes the problem.

 

First do a Windows Chkdsk if there is data corruption or bad sector Windows will attempt to relocate data and mark it as bad sector.
Since you are on Samsung Disk
Follow instruction and download this.
It should be able to fix the bad sectors.

 

Uggh, amateurs!!

I should take a baseball bat to all those who advise wiping the drive or getting rid of it.

If I got rid of every drive I had with bad sectors, I'd be broke. People don't realize that harddrives get bad sectors during the normal course of operation. Normally, they aren't reported to the OS and the HD controller just remaps the sector.

You shouldn't be so much worried about the bad sector count as much as actual SMART errors. If you receive one of those, then your drive is really screwed.

 

Well, uh...

I was about to disagree with you, then see where you're going....but I think your posts can lead to greater confusion.

Bad sectors can be and are something to worry about--they are just not a reason to wipe out a drive or throw it out.

When you look at a SMART report or Crystal HDInfo report everything is just as you say. Harddrives all come with additional space on them that are not reported and the drive hardware works in the background to move data from "bad" sectors into those reserved sectors. Most users never know this occurs unless they use some disk reporting tool since as you point out the OS is not even informed.

However, once the OS starts reporting bad sectors, in event viewer or during a chkdsk, your drive has reached a point where the reserve has been exhausted (usually) and bad sectors are reaching a level of concern point.

You absolutely should run a chkdsk /r on such a drive so the OS can mark the sectors as bad and note not to use that area.

This still does not necessarily mean you should replace the drive. I, too, have had dozens of drives that worked flawlessly for years after such "bad" sectors appear. I have an office machine with a bad 4kb block that has not changed in 18 months.

I have also had drives cascade into oblivion once they appear.

Once I find a drive with bad sectors I will simply check it during monthly maintenance to ensure the situation is isolated rather than growing.

 

Correct, but the actual fix is to use the maufacture utility to write the damaged sector off to abstract the information about bad sectors from the OS.
If you only have a few bad sectors it is still ok
However 40% bad sector means time to replace the drive.
Reason is bad sector will cause latency waits as the drive attempt to skip the bad sectors too much will impair drive performance.
I don't think the drive is so smart as to automatic replace sectors, this is usually done when you run the manufacturer drive utility(pardon the smart pun).

 

I do not understand what you mean by this.

That is not correct. Modern drives do reallocate sectors into the "bad pool" when it has difficulty accessing them. Manufacturer drive utilities can be used for low level formatting and communicating with the SMART circuitry, but they are no more effective than the drive's automatic ability to mark bad sectors.

Also, I cannot help but feel this drive is not as bad as it appears. That is why I wanted him to run a chkdsk /r to get a feel for things. I have no idea how accurate crystal hd info is.

 

You need to get with the time man! These days, we got text messaging, car engines which automatically correct itself, tanks which automatically take into account the warping of the main gun barrel due to repeated firings, and Japanese robots.

Seriously. The OS is smart enough to correct bad sectors when it encounters them. It is only a matter of inductive reasoning that a disk controller can do that. Just look at what SSD disk controllers are capable of...

 

What I mean is if you run the manufacturer tool over the disk, the next time you run Windows chkdsk the "bad sectors" will disappear because it has been added to the g-list so Windows don't need to keep track of it.
I had bad sectors a few kb(less than 10 I forgot it is kb or b but it is very small).
After I run the manufacturer tool in DOS mode, the next time I run Windows chkdsk it no longer reports any bad sectors.

Anyway I tried to "get on with the times" as advised and googled.
I can't find anything on auto sector relocation can someone point me to the right direction?

Reason why my inductive reasoning tells me Windows can't do it because harddisk manufacturer implement their "SMARTNESS" in many different versions and each version varies from another.
You can't expect Microsoft to work and keep the whole list into one OS?
Therefore disk manufacturer has to implement the feature in their tool.

 

We didn't say that Windows tells the drive to do it...the process is built into the firmware of the drive.

As for your other assertion....what tool from what manufacturer does that? Because running a tool in DOS or whatever mini-OS the tool uses will not remove a bad sector from Windows. Windows records the bad cluster data in the MFT. No DOS based tool I know of can read the MFT let alone edit and remove it.

Now, VISTA and Windows 7 has an undocumented switch that will purge the bad sector information from the MFT, but if the sector is truly bad, the next chkdsk will simply re-mark the sector is bad. However, it is useful if you can successfully image the disk, the run the chkdsk with the undocumented switch to report the entire (new) disk as clean again.

 

Erm I used the Western Digital CD DOS Tool
Download from Westen Digital Site
Seagate has their own called SeaTools also downloadable from their site
And it works.
Always download and run the DOS version because Windows version run on Windows API which is located on System Disk which you want to scan hence limiting its functionality.
(A surgeon can't do operation on himself)
The DOS version runs like a Linux Live CD from the RAM hence it is free to manipulate the harddisk.

The G-List is data stored in the harddrive detailing which sectors are available etc not the data sectors.
When a harddisk is manufactured it is scanned for bad sectors as well.
These are put into the P-List it is also on the harddisk itself.
G-List (Growth List when sectors dies)
P-List (Permenent List bad sectors when manufacured)
Anyway if you have a bad sector disk try it then tell me.
No need to argue over something when you can just try it out.
I promise it won't murder your harddisk, however you might want to backup your data first.

I also think logically the drive firmware cannot do that.
Because when it is scanning the whole drive for that auto relocation function
it cannot do anything else because it is busy.
If the drive does that it may cause random performance degradation.

 

It's not scanning the whole drive. It relocates the sector on the fly. Hence the Relocate sector SMART attribute.`

 

First of all, I am not arguing with you. I have used Datalifeguard tools many times over the years and cannot recall the function of which you speak. Granted, I cannot recall the last time I used the DOS version, but I wanted to know if it was indeed the DataLifeguard Tool we were talking about, (or Seatools, or whatever). I should also note, that your description of a surgeon operating on himself is not appropriate if you have a multiple drive system--ie, Windows is loaded on disk 0, and the drive you are checking is disk 1.

I attempted to download the Dos Datalifeguard tools, but like thousands of others I ran straight into the "unable to locate the License Agreement file notice that prevents most people from using the Dos tool.

I looked through the text file (the basic instructions) and find no mention of the function of which you speak. I loaded the windows version and find no mention of the function of which you speak--unless you are just referring to the quick test or extended test.

Either way, what you are saying is simply not making much sense. You claim the only way to reallocate the sectors is to use the manufacturer tool, but if you view any disk information tool you can clearly see the reallocation sector count. If the drive is not doing it automatically, what, pray tell us, is (assuming no one has run your DOS tool)?

Furthermore, your DOS tool, which is running Caldara DOS, cannot read the MFT which is where the bad cluster information is stored. While it is certainly possible the datalifeguard tool can mark sectors as suspect and move them to the G-List before Windows marks them as bad, I am not denying that, such an act won't have an impact on the MFT table unless one runs chkdsk /b within Vista or Windows 7, which purges the bad sector info from the MFT.

As for logically not believing the firmware can mark sector bad or not, again, the proof is staring you in the face when you see the sector reallocation metric in SMART. Sectors ARE being moved there. It's not Windows doing it (these sectors are marked supect before Windows ever counts a sector as bad), it is not DataLifeguard (most people never even run it.

You're simply not making sense. You know a good bit about how harddrives worked (or you read a post somewhere and just regurgitated some facts), but you are not thinking this through.

Or, you are simply not making yourself clear.



Second, the evidence is right in front of you. The drive

 

OK, never mind about the arguments, I will walk you through the Running of the DOS Tool.
1)Grab an Empty CD-R
2)Download the DOS Tool
Data Lifeguard Diagnostic for DOS (CD)
3)Burn to CD as an Image (Lowest Speed is Better)
4)Put it into the tray, go into BIOS set it as primary boot
5) I encountered this issue on my laptop, it doesn't boot from it instead it hung.
It has been quiet a while before WD updated the Tool.
Also remove all other drives you don't want to scan it with, it is a very primitive software so I do not know other effects.
Things I suspect that result in this is the new AHCI mode or maybe southbridge updates.
So what I did was I took an external USB DVD-Drive put it as primary boot and booted from it.
Then run it as usual.
Press Esc at the license agreement.
Press Y then Enter.
Select Extended Test then go for a cup of coffee.
To double confirm running it the second time should yield no errors.
You have to use WD Tool for WD Drive only, SeaTools for Seagate Drive
The reason the Windows Tool has limited functionality is because it run Windows API that is the main reason even if you are doing it on Drive 2 Windows already mounted the drive so there is limitation on what it can do.
When you do extend test the tool will automatically fix all issues.

And as for moved sectors are you sure it is not referring to the P-List? What the manufacturer redirected during manufacturing detection of bad sectors?
Manufacturing is never perfect bad sector are already produced so they need to scan for them and add the bad sectors to the P-List prior to shipping.

What I did was I reformated after using the tool, was going to upgrade to Win7 anyway.
So IF the Tool did nothing as you said there should be bad sectors detected however on my new installation Zero bad sectors were detected when a fresh chkdsk was done.

 

That is what I was getting at--you just used the extended scan. I thought there was some feature in the DOS tool I was unaware of.

We are talking apples and oranges here. One works had the hardware level and one works at the software level.

The extended scan will indeed look for suspect sectors and reallocate them to the don't use pool and simultaneously take a sector from the unallocated sector pool to replace them.

All drives ship with some extra space that is used for this purpose. Sectors go bad. Your drive will do this automatically as it encounters bad sectors, but the tool will jump start the process. I don't know if it will do a more aggressive search, but it will also scan unused areas of the harddrive.

The Windows tool will do the same thing.

They do not fix bad sectors. They simply mark them as not to be used, and then take some space from the unused pool so your harddrive appears complete.

In addition to the harddrive automatically (or as directed with the datalifeguard tool) finding bad sectors, Windows will also find and mark sectors as bad. When it finds a sector it is having difficulty reading or writing to, it will simply mark the sector as bad and remember not to use it again. Sometimes, Windows makes a mistake and a sector is not really bad. Sometimes it marks it bad before the harddrive marks it bad.

Either way, once windows marks a sector bad, it places an entry in the MFT so it will not be used. If you run chkdsk, it will report xx bytes in bad sectors.

There is only one way to remove the bad sector information from the MFT on a working system and that is to run chkdsk with the undocumented /b switch. You can even watch it do it. Somewhere during the run, it will say something along the lines of purgining bad sector information, chkdsk then goes on to complete the scan, checking each sector again, and if it finds it to be suspect, it remarks it as bad. It does not, however, tell the drive to reallocate the bad sector into its bad pool. Windows simply says "I am not going to use it."

I suspect if Windows does not ever check that sector again, it is possible the drive will not reallocate the sector since the drive can only determine a sector is bad by looking at it. Purely speculation on my part.

Datalifeguard cannot access the MFT, but now that you have explained what you did, you stumbled on the other way...YOU FORMATTED YOUR DRIVE WIPING OUT THE MFT.

I can not speak directly to your experience. Perhaps Windows marked the sector bad erroneously--I have had this happen. Perhaps your firmware is not properly reallocating sectors and you were compelled to use datalifeguard to reallocate the sectors. I don't know. Or Windows marked it bad before the drive did. But I can tell you are dealing with only partial information.

You can watch a reallocation count grow on a drive over the years without ever running datalifeguard or other tools. Once the extra space is all used up, though, no reallocation takes place and the drive's bad spots grow--sometimes slowly, sometimes exponentially.

I do not throw out a drive when it reports a bad sector, but I do monitor it, or move it to a less critical use.

 

What I meant is the drive don't do it automatically as you said.
You need to run the tool to do sector relocation.
I know it can do sector relocation when you run the tool.
The point is the tool does sector relocation at a Hardware Level while Windows does it in the MFT only.
It is different.
I have tried the Windows Version.
It failed with some "Unable to fix" error message although it was able to detect because the drive is mounted.
I did remove my mobile drive and attach to the desktop to run the Windows Version and it failed with the same message.
In both cases Windows mounted the drive this could be the reason why it failed.
If sector relocation is done on the fly it could easily cause errors because the drive need to consolidate which areas are usable and which are not.
The original P-list which is also sector relocation done at the factory before it is sold could be what you have mistakenly seen as "relocation on the fly".

 

First, I said all that. I said windows marks it in the MFT and you insist the running the DOS tool removes it based on your experience after you deleted the MFT in a reformat.

Second, I said the tool can jump start the process by scanning the drive.

I will allow that the windows version may not have worked for you--personally, I have no need to run the tool for this purpose because...

The drive DOES do it automatically. You DO NOT need to run the tool. If you wish to, hey, knock yourself out.

But it is common knowledge, and whether it feels right to you matters not a lick. The P list refers to the factory allocation; the g list means grown defect list and this process occurs automatically.

http://en.wikipedia.org/wiki/S.M.A.R.T.
http://en.wikipedia.org/wiki/Bad_sector
http://www.t13.org/Documents/UploadedDocuments/docs2006/D1699r3f-ATA8-ACS.pdf
http://www.mjmdatarecovery.co.uk/data-recovery-articles/bad-sector-errors.html
http://www.patentstorm.us/patents/5917724/description.html

Even the Sea Tools (the program you have mentioned several times) reference material specifically states

...

<blockquote>By design, modern disc drives maintain spare sectors for reallocation purposes. Usually, sectors become
difficult to read long before they become impossible to read. In this situation the actual data bytes in the
sector are preserved and transferred to the new spare during a sector reallocation. Similarly, when a disc
drive writes data and encounters a problem, the drive firmware retires the problem sector and activates a
replacement before giving successful write status.</blockquote>

I can find thousands or references to this automatic process and all you have is "it just doesn't feel right to you?"

You said earlier there is no point arguing about something and I can simply download the tool. Rather than stubbornly clinging to this bizarre notion, and arguing with everyone else, why not spend a little time with google?