Bad sectors on a new drive

[Folkert Rienstra]

It would mean the OS (and user) would see a bad sector that might
otherwise be taken care of by the drive.

And what good would that do?

(Almost) All drives have
bad sectors, and there's nothing wrong with having a few, but usually
on modern drives they're transparently substituted out.

Yes, and?

It's when there are so many that they overwhelm the spares, or when
they grow rapidly, that there's a problem.

And there won't be if you disable that mechanism.
Hmm, why haven't I ever thought of that. That's mind boglingly brilliant.

It's a shame such steps are necessary.

And don't work.

 

[Ron Reaugh]

CJT, (e-mail address removed) escribió en el mensaje,

It could have been cabling or the PS that caused the bad cluster in the
first place -- now that it exists, of course you'll see it every time
until you reformat. The drive's bad sector management won't see it
unless the software actually tries to read the cluster, and I think
having it marked as bad will keep that from happening.

I used Seatools to write zeros (full) to the disk, and then formated it again.
The bad clusters were still found, starting with cluster 308592.

If Seatools found the drive to be ok then likely something else other than
the drive is causing the problem.

Can you CONFIRM that your mobo's BIOS supports drives over about 80GB?

Make that HD the only drive in the system and on the cable as primary master
at the end of the cable and try a clean install of W98SE or something and
see how that behaves.

I did try the "ScanDisk Errors on IDE Hard Disks Larger Than 32 GB" fix at

http://support.microsoft.com/default.aspx?scid=kb;en-us;243450 but the
problem remains.

 

[CJT]

And what good would that do?

No good at all, but it could account for what he's seeing if he had done it.

Yes, and?




And there won't be if you disable that mechanism.

Huh? Nobody said that (except you).

 

[news.ntlworld.com]

Ron Reaugh, (e-mail address removed) escribió en el mensaje, <ACyub.82170$Ec1.4141962@bgtnsc05-
news.ops.worldnet.att.net>, Wed, 19 Nov 2003 00:15:28 GMT

If Seatools found the drive to be ok then likely something else other than
the drive is causing the problem.

Can you CONFIRM that your mobo's BIOS supports drives over about 80GB?

No The BIOS (Award Software International, Inc. 4.51 PG 02/12/99) does not support large drives.
I had to set the drive (using Seagate Disk wizard) to report a lower size (10Gb)before the BIOS
would recognise it.
Seagate Disk wizard installed Drive overlay (DDO) on the drive but W98 would not load. There was a
BSOD caused by "tdimsys.vxd". When I removed the DDO windows loaded as normal.
Explorer shows the drive's full capacity of 111Gb without DDO.

Make that HD the only drive in the system and on the cable as primary master
at the end of the cable and try a clean install of W98SE or something and
see how that behaves.

I may try that later.

 

[Svend Olaf Mikkelsen]

By benign, did you mean that bullshit where they try to shove off their
own stupidity on to Phoenix for offering different CHS translations
(that is limited to 8GB) when that obviously can't have anything to do
with problems that concern addressing above 32 GB?

I mean that they mention that Scandisk reports errors, but not that
all data which should be written more than 32 GB into the disk are
written at the beginning of the disk.

 

[news.ntlworld.com]

news.ntlworld.com, (e-mail address removed) escribió en el mensaje,

Ron Reaugh, (e-mail address removed) escribió en el mensaje, <ACyub.82170$Ec1.4141962@bgtnsc05-
news.ops.worldnet.att.net>, Wed, 19 Nov 2003 00:15:28 GMT

No The BIOS (Award Software International, Inc. 4.51 PG 02/12/99) does not support large drives.
I had to set the drive (using Seagate Disk wizard) to report a lower size(10Gb)before the BIOS
would recognise it.
Seagate Disk wizard installed Drive overlay (DDO) on the drive but W98 would not load. There was a
BSOD caused by "tdimsys.vxd". When I removed the DDO windows loaded as normal.
Explorer shows the drive's full capacity of 111Gb without DDO.

Problem solved. DDO needed to be loaded.
I managed to get windows to load with the DDO on the drive by removing the
following line from config.sys:
device=c:\cdrom\gscdrom.sys /d:mscd000.
Don't know why that was causing a problem.

Thanks to all for helping me solve this problem.

 

[Folkert Rienstra]

I mean that they mention that Scandisk reports errors, but not that
all data which should be written more than 32 GB into the disk are
written at the beginning of the disk.

Doesn't sound like these are related then, other then the number
32GB featuring in both.

 

[Folkert Rienstra]

Sounds like we are on complety different paths.
I'll retry in the other post.

 

[Folkert Rienstra]

Ok, let's redo that.


I took that as:
"I suppose you can also turn the drive's bad sector management off".

Apparently, judging from your later responses, and from another sentence
that similarly didn't make sense either, you probably meant to say:

" I suppose you may have turned off the drive's bad sector management "

It would mean the OS (and user) would see a bad sector that might
otherwise be taken care of by the drive.

Not with unrecoverable read error bad sectors. These will not be automa-
tically 'taken care of' by the drive. Only recoverable read error bad sec-
tors are. Unrecoverable read bad sectors are only taken care of on writes.

Bad sector management is not actually about managing bad sectors but
about managing 'soon to be' bad sectors, before they actually become
unrecoverable bad sectors.

So, in both cases the result would have been the same.

(Almost) All drives have
bad sectors, and there's nothing wrong with having a few, but
usually on modern drives they're transparently substituted out.

That's the plan.
Unfortunately the drive cannot distinguish between physical bad sectors
and bad sector writes where the write has not been completed in a power
failure or where a write went badly due to too high a drive temperature.

It's when there are so many that they overwhelm the spares, or when
they grow rapidly, that there's a problem.

That's correct.

 

[Svend Olaf Mikkelsen]

Doesn't sound like these are related then, other then the number
32GB featuring in both.

The update suggested on the page solves the 32 GB wrap problem.

 

[Folkert Rienstra]

The update suggested on the page solves the 32 GB wrap problem.

Solves *also*, I presume.

So your suggestion is that, although they don't appear to be related,
they are one and the same problem, then?