Im getting an error stating my HD has many bad sectors.
SMART data shows it to be a reallocated sector count error.
The laptop was bought second hand about 6 weeks ago and this is the
first major error so far. After looking around on the web, it seems the disk may be
failing already.
Is this the case or is it possible the sectors are not being checked correctly?
Any help you could give is much appreciated.
On 11/27/2011 06:16 PM, pistolized wrote:
>
> Im getting an error stating my HD has many bad sectors.
> SMART data shows it to be a reallocated sector count error.
> The laptop was bought second hand about 6 weeks ago and this is the
> first major error so far. After looking around on the web, it seems the
> disk may be
> failing already.
> Is this the case or is it possible the sectors are not being checked
> correctly?
>
> Any help you could give is much appreciated.
>
>
while waiting for an authoritative answer i would suggest you plan for
it to die sooner rather than later…by that i mean if you have
anything on that drive which you do not want to loose then now is the
time to
boot from a live cd
mount that drive read only (minimizes potential data loss by writing
to the file in any way (even a log file being written in the background
could cause trouble)…
-use whatever tools you are most familiar with to copy all “don’t
loose” stuff to an off machine location (USB drive, CD, DVD,
floppies…whatever)
do some on line shopping for a replacement disk while you check back
here from time to time to see if someone else has a better idea…
on the other hand: do you know if those sectors were bad before you
bought the device and the number of bad sectors are not increasing? if
so, then you might be safe to postpone step one a few minutes…or
maybe even hours…but, i wouldn’t risk it.
ymmv and i encourage you to also consider the other views which may come
here.
oh! you might check with the drive makers web site and see what kind of
disk inspection and health checking programs are there (AFTER you have
saved all important stuff to a known good drive)…and, also check the
manufacturer’s warranty…maybe it is still covered and they will send
you a new one…or, maybe the warranty ran out some months ago–in which
case i would advise you to STOP reading now, and go to step one.
–
DD http://tinyurl.com/DD-Caveat
openSUSE®, the “German Engineered Automobiles” of operating systems!
Thanks for your speedy reply.
Important data has been copied to an external drive.
I’m going to contact the seller tomorrow if this is likely to develop
into a a full blown disk failure in the near future.
Ive just scanned and attempted recovery of bad sectors on
my Windows partition. No errors were found and the disk was
in usable condition. Does this imply that only the windows partition was
checked or that only openSUSE is detecting the errors? Does this make
it more likely the bad sectors are software based? I may just reinstall
the OS now that data has been backed up, and see what happens.
I have no idea if the sectors were bad before purchase but I would imagine not as the laptop
was bought from a reputable enough local refurb store.
Thanks again.
On 11/27/2011 07:16 PM, pistolized wrote:
>
> Ive just scanned and attempted recovery of bad sectors on
> my Windows partition. No errors were found and the disk was
> in usable condition. Does this imply that only the windows partition
> was checked
if you checked the disk with any disk checking program running in
Windows then you can be sure that only the windows partions were
checked, because no Windows can even see any Linux partitions…to
Windows it just looks like confusing space that it can’t read and
doesn’t understand…
> or that only openSUSE is detecting the errors?
you said the bad sector count came from SMART…read up on S.M.A.R.T.
<Self-Monitoring, Analysis and Reporting Technology
http://en.wikipedia.org/wiki/S.M.A.R.T.> and you’ll know the info comes
not from Linux, but directly from the self-montoring electronics on
the disk itself!
so, i have no idea if the Windows software you ran has the ability to
see the same problems being seen by SMART, but i am sure it didn’t
check the Linux partitions and Linux didn’t say the disk has bad
sectors, instead the disk said that…
NOW: it is possible that SMART is in error and there are zero or very
few bad sectors…
also, it is possible that the disk had all of those bad spots the day it
was born, and none have gone bad since–which is why i ask about what
it had…or, in other words if you see more and more disk sectors being
reported as bad, you can be sure it is getting worse and it may
suddenly fail completely…
well, disk failure is not a matter of if but, rather when…your
disk will fail, my disks will fail–plan on it!
> Does this make
> it more likely the bad sectors are software based?
bad disk sectors are on the disk—not in software…
now, it is true that faulty software can/may/will report wrong data.
> I may just reinstall
> the OS now that data has been backed up, and see what happens.
waste of time…if the SMART software is reporting wrongly it will
report wrongly to every operating system software capable of reading the
data generated. (and, if you find a SMART client which runs in Windows
(rather than the scanning and defragmentation software included in
Windows) you should see exactly the same data being reported by SMART.
–
DD http://tinyurl.com/DD-Caveat
openSUSE®, the “German Engineered Automobiles” of operating systems!
After looking around on the web, it seems even a single bad sector usually indicates a majour problem.
Most people seem to replace their drive after 7 or 8 sectors going bad and experience a majour decline in
disk efficiency and speed. I cant understand how I could have over 65000 genuinely bad sectors and not
experience any adverse effects other than the SMART warning.
Anyway, last 2 questions I have is, how long should I wait before comparing the bad sector count again?
I assume if there is a genuine problem, the disk will become more and more damaged as it’s used.
If bad sectors are in fact increasing I will replace the drive asap.
And also, if the sectors were in fact bad after the birth of the drive, why would a warning be displayed
at a seemingly random time?
Thanks for all your help.
For some years now HDs have an area reserved to substitute bad blocks, which is done automatically by the drive’s firmware. Many (if not all) disk platters that come out of the foundry have defective sectors, that are tested at the factory and remapped to the reserved area.
I know that if you start getting read error messages during operation of the drive it means this reserved area is full and the firmware can’t remap any more sectors. You should immediately:
What I don’t know is if smart is reporting is the total number of bad blocks (not too worrisome) or only the ones that can’t be relocated (time to panic). I think it is the first, but 65000 sectors looks suspiciously lie a 64K pointer limit, so I’d start thinking about replacing the HD anyway.
You’ll probably get more info on the drive manufacturer’s site, specially pertaining to how many bad sectors the drive can cope with.
On 11/27/2011 09:26 PM, pistolized wrote:
> Anyway, last 2 questions I have is, how long should I wait before
> comparing the bad sector count again?
it takes (what?) 49 seconds to check with SMART?
so, do it every day, why not? twice a day…if there are no changed in
the first few days, fall back to once every two days…then once a
week…if after checking every other day it hasn’t grown in (say) a
month…then, once a month, then . . .
> And also, if the sectors were in fact bad after the birth of the drive,
> why would a warning be displayed
> at a seemingly random time?
hmmmm…from your earlier post i assume you don’t have a SMART client
installed (and, used instead a Windows disk inspection software which (i
guess) is not working with SMART) so, the SMART warning couldn’t be
displayed until the SMART software was installed and run, in Linux…
do not overlook brunomcl’s input…as far as i know everything he says
is spot on…oh, just let me add one thing: the number of bad sectors
expected by the manufacture is directly related to the size of the drive…
maybe you have a drive about to go bang, and maybe not…watch for
more bad sectors…if there are none just keep a good backup of all
data and keep checking . . .
and, like we say here: Have a lot of fun!
–
DD http://tinyurl.com/DD-Caveat
openSUSE®, the “German Engineered Automobiles” of operating systems!
Well the SMART client was the first and only utility to give any warning thus far.
SMART software came standard with the openSUSE 11.4 distro.
Anyway, following both of your recommendations and finding the receipt
will allow for return of the laptop within the next two weeks I’ve decided to
contact the seller. I just hope he has a spare drive lying around and wont have to
replace the whole machine.
Thanks a lot guys.
> Thanks a lot guys.
welcome, let us know how it goes…
–
DD
openSUSE®, the “German Engineered Automobiles” of operating systems!
Before contacting the vendor, I decided to try the manufacturers website.
The drive is an ATA-7: Hitachi HTS541612J9SA00. The Hitachi site provides a
drive fitness test which I download before burning the image to a disk. The test can be run
in a quick or advanced mode. I ran both. Both check, among other things, the status of the
SMART diagnostics utility and returned a disposition code ‘0x00’ meaning there are no errors.
The confusion continues!
So, I have deleted my openSUSE and swap partitions.
Extended my Windows partition to the max and installed HD Tune.
HD Tune highlights the reallocated sector count and shows the following information:
Current: 100, Worst: 100, Threshold: 5, Data: 65541 Status: OK
So, it shows the same information as shown by SMART when it was running in Linux
but, states that it is OK and there are no health issues.
On 11/28/2011 05:56 PM, pistolized wrote:
>
> it is OK and there are no health issues.
so, perhaps the subject line of this thread “Reallocated sector count
Value: 65541 sectors” is not the same as what you wrote in your opening
line “an error stating my HD has many bad sectors”??
or, maybe “reallocated sector” are not “bad sectors”??
i really don’t know the answer to those questions–but, happy to see the
confusion wasn’t due to a Linux problem!! 
–
DD
openSUSE®, the “German Engineered Automobiles” of operating systems!
On 2011-11-28 17:56, pistolized wrote:
> So, it shows the same information as shown by SMART when it was running
> in Linux
> but, states that it is OK and there are no health issues.
SMART will not signal “not ok” till all spare sectors are mapped. If the
number does not increase, you can use the disk (if young). Another of the
lines will tell you how many thousands hours the disk has.
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
Usually it’s easy to open the drive enclosure under the laptop. If you feel confident you can do it and it has no impact in whatever warranty you may have, I’d recommend you take a look at the drive manufacturing date stamped on the label. IME desktop HDs last about 5 years, a laptop’s may last less (smaller, hotter, etc.).
On 2011-11-29 15:46, brunomcl wrote:
>
> Usually it’s easy to open the drive enclosure under the laptop. If you
> feel confident you can do it and it has no impact in whatever warranty
> you may have, I’d recommend you take a look at the drive manufacturing
> date stamped on the label. IME desktop HDs last about 5 years, a
> laptop’s may last less (smaller, hotter, etc.).
There is no need to look at the label to know how many hours it has been
running. Just use smartctl.
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
Not how many hours, but age. Of course your approach works too if the OP knows the drive MTBF - supposing you can trust the vendor spec, which I don’t. I’ve seen too many deviations (for less) to give much credence to it.
But most drives I’ve used tend to fail in or after 5 years from the stamped date.
Right well I took it back to the vendor this morning. He ran a number of checks(wiping the disk in the process)
and found no errors and that the disk is in usable condition so he wont replace it.
He advised me not to install two operating systems strangely enough…I dont see how that would matter.
But, other than the constant warning displayed by SUSE I’m yet to have any performance issues.
Right! But, only SUSE displays the “not ok” message. HD Tune in Windows displays the same results
with an “ok” status. So, considering there may be no scenario of impending doom…
How do I stop the SMART warnings in SUSE popping up all the time?
By the way, the reallocated sector count has not increased at all since the error messages began.
Checked the drive just there, it was manufactured in July of 2007. I haven’t reinstalled SUSE yet as I’m busy with assignments, I’ll do it tomorrow evening so I can check how many hours left the drive supposedly has.
Thanks for all your replies.
Oh I should clarify when I say “only SUSE displays the “not ok” message” I mean using SMART in SUSE.
And HD Tune in Windows uses SMART also. Plus the Hitachi disk check I ran displayed no errors even though
that uses SMART also!
Hard to make sense of :\
On 2011-12-01 00:46, brunomcl wrote:
> But most drives I’ve used tend to fail in or after 5 years from the
> stamped date.
It may be less than two years old and have 10 thousand hours of use. That’s
a lot.
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
On 2011-12-01 23:46, pistolized wrote:
>
> Right well I took it back to the vendor this morning. He ran a number of
> checks(wiping the disk in the process)
> and found no errors and that the disk is in usable condition so he wont
> replace it.
Notice that bad sectors are remapped and that the typical windows disk
check, even a surface check, will see nothing at all.
Post the complete output of “smartctl -a /dev/sda” here between code tags
(advanced editor, # button)
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)