what are bad blocks exactly?

  • Thread starter Thread starter Ni©
  • Start date Start date
N

Ni©

Hi,

I get an error message in event viewer saying that the disk has bad blocks.
Also in disk management, there is an exclamation mark and although the
status is healthy it also says it's at risk.
Is this error a physical problem?
I tried already to run checkdisk with that fix option, still the same.

The disk is a maxtor 80Gb IDE disk and only 6 months old.
Is a acceptable that a disk has - if they are physical - errors that soon or
can I just demand to replace it with a new one?

Thanks a lot in advance,
 
Hi,

I get an error message in event viewer saying that the disk has bad blocks.
Also in disk management, there is an exclamation mark and although the
status is healthy it also says it's at risk.
Is this error a physical problem?
I tried already to run checkdisk with that fix option, still the same.

The disk is a maxtor 80Gb IDE disk and only 6 months old.
Is a acceptable that a disk has - if they are physical - errors that soon or
can I just demand to replace it with a new one?

Thanks a lot in advance,
Click to expand...

Bad blocks are blocks on the disk drive that don't read or write reliably.
They normally indicate a physical problem with the drive, although some
software problems and virus can make it look like there are bad blocks. If
there are are no virus (you have checked I hope), and chkdsk will not fix
the problem, the drive is probably bad. Run Maxtors utility to be sure, but
you probably need to exchange the drive.

JT
 
Hi JT,

Sorry for this late reply but I still have to try that maxtor utility.
Thanks a lot already for that suggestion!

--

Ni©


"When you lose, don't lose the lesson."
 
Dunno if you have to exchange the drive... I always thought that
checkdisk would identify the bad blocks and add them to a chain that
looks like a used file i.e. make the block look like it's already
allocated so it wouldn't be used for real data.

Of course, I've been wrong before...

:o)
 
Dunno if you have to exchange the drive... I always thought that
checkdisk would identify the bad blocks and add them to a chain that
looks like a used file i.e. make the block look like it's already
allocated so it wouldn't be used for real data.

Of course, I've been wrong before...

:o)
Click to expand...

On modern drives, with a number of hidden spare blocks to replace failed
blocks, if you have a drive that has bad blocks even after running the
manufacturers diagnostics which should remap the bad block, the drive is
failing and should be replaced. Sometimes the drive might last a long
before failing completely, but the odds are against that, so if it under
warranty, why take a chance, and if it is not under warranty, how
important is your data?

JT
 
I find this a bit ironic as drives, even going back to the 1970's, had bad
blocks and you mapped them out and the drives were reliable. It wasn't a
sign that the drive was going bad, it was a manufacturing defect (like bad
pixels on LCD displays). Is some other mechanism causing the problem these
days or are people mistakenly considering head "touch-downs" as bad blocks?
(When the heads start touching down, that IS the beginning of the end
usually caused by some contaminant in the drive...which could even be some
oxide itself that has flaked off or been knocked off by a head touch-down
caused by lack of air flow, power failure, etc.) The PC drives that I've
had fail haven't exhibited this trait...
 
In the early days of MFM & RLL, there were no spare blocks, so all bad
blocks showed up and had to be marked as such by the OS. Once drives got
bigger and smarter with EIDE, a percentage of the drive as set aside as
spare blocks, and any bad block from the factory was automatically remapped
to a good one from the spare blocks pool, so there were no bad block from
the factory. Smart and the manufacturers diagnostics will automatically
remap questionable blocks unless there are no more spares. With modern
drives and about 1% of the drive in the spare pool (varies by
manufacturer), once the spare blocks run out, you have a lot of bad spots
on the drive, and the odds are they are going to get worse. Even in the
days of MFM & RLL, bad blocks showing up that weren't there before was a
cause for concern. You ran utilities like Spinrite or Norton Disk Doctor to
rigorously test the drive to see if it was just a one time thing.

On a modern EIDE drive, if bad blocks are showing up during the warranty
period, replace it. If outside of the warranty period, be very wary.

JT
 
JT said:
You ran
utilities like Spinrite or Norton Disk Doctor to rigorously test the
drive to see if it was just a one time thing.
Click to expand...

Caution: "Rigorously" is the give-away here. Be sure to back up all
irreplacable data. Disk Doctor has been known to destroy marginal disks.
 
Caution: "Rigorously" is the give-away here. Be sure to back up all
irreplacable data. Disk Doctor has been known to destroy marginal disks.
Click to expand...

Agreed. Even today, when you are going to do thorough testing on a disk
drive, backing up all the data first is a great idea.

JT
 
Thanks for updating me. Do you know what the failure mechanism is with this
situation of ever-increasing bad blocks?

Also, what you state makes it really sound like buying a drive is a major
crapshoot vs. drives in the past. Previously, you KNEW how much of the
surface was bad and now that data is masked (i.e., the drive you buy could
be using 0% of the spare blocks pool or 100% of the spare blocks pool and
you wouldn't know the difference).
 
Failure mechanism is pretty much what it has always been. Surface damage.
Mechanical problems such as bearings, etc.. Corrosion, contamination.
Degradation of electronics, amplifiers, heads, etc..

The amount of the drives spare blocks used can often be found by the drive
diagnostics or smart utilities. Because of the improvements in tracking and
bonding of surface coatings, that should stay pretty much constant for the
life of the drive. It is not like the early days of stepper motors where
heads would not always be centered on the track. Using voice coils and
embedded positioning information, it is almost impossible for the head not
to be over the track. Also, with the differences in how the surface is
made, coating coming off is extremely rare. With the ever tighter
tolerances and closer head spacing, any surface irregularity is both rare
and critical.

That is why it is rare to see real bad blocks show up anymore. Drives
will often fail before any new bad blocks show up now, even checking those
that get remaped by smart, unlike earlier days. In the last 5 years I have
seen few drives fail because of bad blocks. Those that developed bad
blocks that I kept to play with seemed to soon enter a death spiral of more
and more bad blocks, soon becoming unusable. That was a very small
percentage of drives though.

JT
 
Back
Top