Prev: simple two-factor solution for logins to Linux system: one time passcodes
Next: Seeking suggestions for site search
From: Stefan Patric on 10 Oct 2009 12:42
On Sat, 10 Oct 2009 00:55:52 -0700, Al wrote:

> [snip]
> The other day I used badblock on a drive that Partition table doctor had
> told me had a bad sector, but badblocks failed to find it.
>
> This has got me puzzled. Anyone able to shed any light on why this might
> be?

Maybe, Partition Table Doctor was wrong. FYI, badblocks doesn't read
sectors, per se; it reads blocks. Hence, its name. In any case,
badblocks' nondestructive, read only check, which is the default, isn't
foolproof: You may be able to read the data on a block, but the block
won't accept a write. This is rare, but it happens. The destructive,
read-write checks are more thorough and accurate, but slow. Don't ever
use this check, if you are testing a drive and don't want all the data on
the drive destroyed.

Read the man and info files for badblocks to really know how it works.
In a terminal, type:

man badblocks

or

info badblocks


Stef
From: Al on 10 Oct 2009 15:55
On Oct 11, 5:42 am, Stefan Patric <n...(a)thisaddress.com> wrote:
> On Sat, 10 Oct 2009 00:55:52 -0700, Al wrote:
> > [snip]
> > The other day I used badblock on a drive that Partition table doctor had
> > told me had a bad sector, but badblocks failed to find it.
>
> > This has got me puzzled. Anyone able to shed any light on why this might
> > be?
>
> Maybe, Partition Table Doctor was wrong.  FYI, badblocks doesn't read
> sectors, per se; it reads blocks.  Hence, its name.  In any case,
> badblocks' nondestructive, read only check, which is the default, isn't
> foolproof: You may be able to read the data on a block, but the block
> won't accept a write.  This is rare, but it happens.  The destructive,
> read-write checks are more thorough and accurate, but slow.  Don't ever
> use this check, if you are testing a drive and don't want all the data on
> the drive destroyed.
>
> Read the man and info files for badblocks to really know how it works.  
> In a terminal, type:
>
>    man badblocks
>
> or
>
>    info badblocks
>
> Stef

Thanks for that reply. At the moment I have yet to find a good
explination of the difference between a block and a sector, well at
least I can understand.

As someone pointed out a block is not a sector, nevertheless they both
claim to do surface tests.

-Al
From: The Natural Philosopher on 10 Oct 2009 19:29
Al wrote:
> On Oct 11, 5:42 am, Stefan Patric <n...(a)thisaddress.com> wrote:
>> On Sat, 10 Oct 2009 00:55:52 -0700, Al wrote:
>>> [snip]
>>> The other day I used badblock on a drive that Partition table doctor had
>>> told me had a bad sector, but badblocks failed to find it.
>>> This has got me puzzled. Anyone able to shed any light on why this might
>>> be?
>> Maybe, Partition Table Doctor was wrong. FYI, badblocks doesn't read
>> sectors, per se; it reads blocks. Hence, its name. In any case,
>> badblocks' nondestructive, read only check, which is the default, isn't
>> foolproof: You may be able to read the data on a block, but the block
>> won't accept a write. This is rare, but it happens. The destructive,
>> read-write checks are more thorough and accurate, but slow. Don't ever
>> use this check, if you are testing a drive and don't want all the data on
>> the drive destroyed.
>>
>> Read the man and info files for badblocks to really know how it works.
>> In a terminal, type:
>>
>> man badblocks
>>
>> or
>>
>> info badblocks
>>
>> Stef
>
> Thanks for that reply. At the moment I have yet to find a good
> explination of the difference between a block and a sector, well at
> least I can understand.
>
> As someone pointed out a block is not a sector, nevertheless they both
> claim to do surface tests.
>

IIRC the phyical disk is tracks and sectors. However there are rather
too many of those to be listed individually, so its conventional to
aggregate sectors into blocks. Sadly that means if one byte in one
sector in one block is bad, the whole block gets marked bad.

This understanding is some years old, and things may well be different now.

> -Al
From: Stefan Patric on 11 Oct 2009 00:45
On Sat, 10 Oct 2009 12:55:09 -0700, Al wrote:

> On Oct 11, 5:42 am, Stefan Patric <n...(a)thisaddress.com> wrote:
>> On Sat, 10 Oct 2009 00:55:52 -0700, Al wrote:
>> > [snip]
>> > The other day I used badblock on a drive that Partition table doctor
>> > had told me had a bad sector, but badblocks failed to find it.
>>
>> Maybe, Partition Table Doctor was wrong.  FYI, badblocks doesn't read
>> sectors, per se; it reads blocks.  Hence, its name.  In any case,
>> [snip]
>
> Thanks for that reply. At the moment I have yet to find a good
> explination of the difference between a block and a sector, well at
> least I can understand.

Here's a basic explanation of hard drive geometry and addressing.

http://en.wikipedia.org/wiki/Cylinder-head-sector


Stef
From: Douglas Mayne on 11 Oct 2009 12:21
On Sat, 10 Oct 2009 00:55:52 -0700, Al wrote:

> Hi Folks,
>
> I have been getting into the world of Linux, partly because it has
> some quite good tools for dealing with DR. I have been slowly find
> ways to do things that I used to do in Windows in Linux.
>
> One of these is the surface test of a hard disk. In WIndows I used to
> use Partition Table Doctor, which is a farily generic windows based
> surface tester among other things.
>
> The other day I used badblock on a drive that Partition table doctor
> had told me had a bad sector, but badblocks failed to find it.
>
> This has got me puzzled. Anyone able to shed any light on why this
> might be?
>
> Admittadly the drive only had 1 bad block, but shouldnt the results be
> the same?
>
> Cheers in advance,
>
> -Al
>
It is my understanding that some drives have a limited set of unused/spare
sectors which can be used to replace bad sectors. Perhaps
the drive noticed the error and automatically replaced it. Or else, it
could be that the Windows tool performed the remap for you. Another Linux
tool to check out is smartctl. That command could possibly shed
some light on what happened. The manual is avaialable: man smartctl.

The native disc tools may also shed some light. For example, Seagate's
SeaTools can give event history for some Seagate drives. Other
manufacturers have similar tools.

Another possibility is that you could be in an intermediate
failure mode. The reported disk read error did occur at that time but it
is readable now, but the sector will probably fail in the near future.
A disk read error is only reported after a series of failed reads by the
OS.

--
Douglas Mayne
First  |  Prev  |  Next  |  Last
Pages: 1 2 3
Prev: simple two-factor solution for logins to Linux system: one time passcodes
Next: Seeking suggestions for site search