Hello everybody
I'm using a registered version of HD Sentinel to inspect some of our HDs trying to find out whether they're still reliable or not.
I am trying to understand how "bad" is a drive with a single - or just a few - bad sectors on it. In my limited knowledge of the matter I know that every HD has a reserved area on them where the naturally developing bad sectors that come across HD lifespan are stored in a way that the user cannot detect.
So for example if I run scandisk and it finds "no bad sector" it does not mean that some bad sector are not actually present on the disk. On the contrary I know that when scandisk finds some bad sectors it means that the reserved area is full.
What I'm trying to sort out is whether the bad sectors counter in HDSentinel shows any bad sector on the drive or just the ones that fall out of the "hidden" area. So, to make an example, when HDSentinel says there is only 1 bad sector, does it mean that there is actually only ONE bad sector on the drive or does it mean that the reserved area is full and one sector couldn't have been relocated?
Hopefully the question is clear, please forgive me for my bad english and feel free to ask if something is not clear!
Many thanks
Antonio
Bad sectors count
-
hdsentinel
- Site Admin
- Posts: 3115
- Joined: 2008.07.27. 17:00
- Location: Hungary
- Contact:
Re: Bad sectors count
Hi Tony359,
First of all, thanks for using registered version of HD Sentinel.
You are absolutely correct. All modern hard disks have a reserved space which is called spare area. In normal operation, these sectors are not used but if the drive found any problems with the "normal" data sectors, it tries to move the contents to this spare area (this is reallocation), marks the original sector as bad and the drive never use it again. All further reads and writes affecting that sector are redirected to the spare area instead. In that case, the bad sector counter (attribute 5 or 198, depending on the drive) increases.
You are again correct. If the drive has some such bad sectors, scandisk and other software based tests do not find any problems, because they cannot access the original (bad) area. As long as the number of such bad sectors is relatively low (just some or around 10) and the status is stable (no new bad sectors born after longer usage) the drive is fine and can be used for a longer period.
Just a note: when scandisk finds bad sectors, it may happen even when the drive is 100% perfect. Scandisk finds logical problems (issues with the file system only) and file system corruption may occur even on a perfect drive for example upon a power loss, power failure, data cable issues and so.
So an error reported by scandisk may not neccessary mean that the spare area is full. Usually the File -> Reinitialise disk surface test in Hard Disk Sentinel (or even a quicker write test in that menu) may correct the issue and then creating and formatting a new partition would have no such problems.
HD Sentinel reads and displays the value of the bad sector counter which counts all sectors reallocated to the spare area.
So if you see 1 bad sector, it means that the hard disk found only one sector, reallocated it and then use the spare area instead of that sector. The spare area may hold thousands of sectors before it is filled - but if it is filled, it means really big problems with the drive (big scratch, metchanical shock or even a crashed/broken head which cause many thousands or even millions of bad sectors).
This is why the health value is still very high and the status is excellent if you see 1-2 (or 10) bad sectors reported in HD Sentinel.
If this number is high and the status is stable, you can even set the software to ignore these problems and report only further issues by using a negative offset value at the affected S.M.A.R.T. attributes.
Please click on the "?" button next to the text description (which displays the possible problems with the disk) for further information about the bad sectors. If you have further questions or I can help, please let me know.
First of all, thanks for using registered version of HD Sentinel.
You are absolutely correct. All modern hard disks have a reserved space which is called spare area. In normal operation, these sectors are not used but if the drive found any problems with the "normal" data sectors, it tries to move the contents to this spare area (this is reallocation), marks the original sector as bad and the drive never use it again. All further reads and writes affecting that sector are redirected to the spare area instead. In that case, the bad sector counter (attribute 5 or 198, depending on the drive) increases.
You are again correct. If the drive has some such bad sectors, scandisk and other software based tests do not find any problems, because they cannot access the original (bad) area. As long as the number of such bad sectors is relatively low (just some or around 10) and the status is stable (no new bad sectors born after longer usage) the drive is fine and can be used for a longer period.
Just a note: when scandisk finds bad sectors, it may happen even when the drive is 100% perfect. Scandisk finds logical problems (issues with the file system only) and file system corruption may occur even on a perfect drive for example upon a power loss, power failure, data cable issues and so.
So an error reported by scandisk may not neccessary mean that the spare area is full. Usually the File -> Reinitialise disk surface test in Hard Disk Sentinel (or even a quicker write test in that menu) may correct the issue and then creating and formatting a new partition would have no such problems.
HD Sentinel reads and displays the value of the bad sector counter which counts all sectors reallocated to the spare area.
So if you see 1 bad sector, it means that the hard disk found only one sector, reallocated it and then use the spare area instead of that sector. The spare area may hold thousands of sectors before it is filled - but if it is filled, it means really big problems with the drive (big scratch, metchanical shock or even a crashed/broken head which cause many thousands or even millions of bad sectors).
This is why the health value is still very high and the status is excellent if you see 1-2 (or 10) bad sectors reported in HD Sentinel.
If this number is high and the status is stable, you can even set the software to ignore these problems and report only further issues by using a negative offset value at the affected S.M.A.R.T. attributes.
Please click on the "?" button next to the text description (which displays the possible problems with the disk) for further information about the bad sectors. If you have further questions or I can help, please let me know.
Re: Bad sectors count
Thank HDSentinel for your detailed reply.
In your experience, would you consider a drive with a few bad sectors count (<10) not entirely reliable? We ship drives around and if a drive fails we have to ship another one using very expensive shippers to have it to our customer on time.
I have been testing 40 of our drives to make some kind of stats and I found only ONE drive with just ONE bad sector on it. That leads me to think that if I ever encountered a drive with 10 bad sectors, I should consider it as faulty for our needs.
What is your opinion on that? (I would not consider you responsible for any shipped drive found faulty on the field
)
Many Thanks
Antonio
In your experience, would you consider a drive with a few bad sectors count (<10) not entirely reliable? We ship drives around and if a drive fails we have to ship another one using very expensive shippers to have it to our customer on time.
I have been testing 40 of our drives to make some kind of stats and I found only ONE drive with just ONE bad sector on it. That leads me to think that if I ever encountered a drive with 10 bad sectors, I should consider it as faulty for our needs.
What is your opinion on that? (I would not consider you responsible for any shipped drive found faulty on the field
)Many Thanks
Antonio
-
hdsentinel
- Site Admin
- Posts: 3115
- Joined: 2008.07.27. 17:00
- Location: Hungary
- Contact:
Re: Bad sectors count
In most cases, few bad sectors (<10) should not be a problem.
Because of the high density, even a microscopic scratch or dust can make a sector bad on the disk surface. But this is why the spare area is available as then it will be used instead of the original location.
In these cases, Hard Disk Sentinel still displays excellent/very high (90% and above) health levels and also display a warning that
At this point, warranty replacement of the disk is not yet possible, only if the health drops further.
On these drives, using a complete surface analysis (File menu -> Surface tests) is a good idea as it verifies that
- all sectors are reallocated so the available surface is usable
- or are there further sectors which will require reallocation
Problem would occur only if the number of bad sectors would constantly increase. In case of relatively low number of bad sectors I recommend constant monitoring and immediate backup when the health changes (even with only 1 %).
Generally, manufacturers and resellers only need to replace (in terms of warranty) when the health goes down to 0%. This is the point when the number of problems (bad sectors or other kind of problems) reach the error-level threshold set by the manufacturer. However, this is often cannot be reached (for more details, please check http://www.hdsentinel.com/smart )
So I recommend earlier replace: disks with low health (25% or lower) should not be used for data storage because of the numerous problems.
If the drives store critical information and data integrity is not a question, I'd recommend to trust 100% drives only.
Because of the high density, even a microscopic scratch or dust can make a sector bad on the disk surface. But this is why the spare area is available as then it will be used instead of the original location.
In these cases, Hard Disk Sentinel still displays excellent/very high (90% and above) health levels and also display a warning that
At this point, warranty replacement of the disk is not yet possible, only if the health drops further.
On these drives, using a complete surface analysis (File menu -> Surface tests) is a good idea as it verifies that
- all sectors are reallocated so the available surface is usable
- or are there further sectors which will require reallocation
Problem would occur only if the number of bad sectors would constantly increase. In case of relatively low number of bad sectors I recommend constant monitoring and immediate backup when the health changes (even with only 1 %).
Generally, manufacturers and resellers only need to replace (in terms of warranty) when the health goes down to 0%. This is the point when the number of problems (bad sectors or other kind of problems) reach the error-level threshold set by the manufacturer. However, this is often cannot be reached (for more details, please check http://www.hdsentinel.com/smart )
So I recommend earlier replace: disks with low health (25% or lower) should not be used for data storage because of the numerous problems.
If the drives store critical information and data integrity is not a question, I'd recommend to trust 100% drives only.
Re: Bad sectors count
Thank you very much for your help and your time.
Antonio
Antonio