CHIRPING Hard Drive? About to die?

[Bob Brown]

I occasionally hear a CHIRP noise from my drive. It is more frequent
if I do a spyware scan or defrag, but it does happen when the comp is
idle like at night just before falling asleep.

SMART reports no errors or anything out of spec.
I've also downloaded several programs to "test" the drive and no
errors were reported.

I cannot run the WinXp GUI version of SEATOOLS but was able to burn a
cd dos version and run it and it also found zero problems/errors.

The drive is...
http://www.newegg.com/product/product.asp?item=N82E16822148140
Seagate Barracuda 7200.10 ST3320620AS (Perpendicular Recording
Technology) 320GB 7200 RPM 16MB Cache SATA 3.0Gb/s [I have it set to
SATA ONE because the motherboard does not support SATA II

Opinions/suggestions/conjecture welcome.
thanks.

 

[geoff]

I decided to take the plunge and used spinrite, $80, and I found the tools,
like seatools and the ones from western digital are not very good. SR fixed
the problems my drive had and it was a lot more convenient then pulling
it/back it up/sending it somewhere, etc.

-g

 

[Frank McCoy]

I decided to take the plunge and used spinrite, $80, and I found the tools,
like seatools and the ones from western digital are not very good. SR fixed
the problems my drive had and it was a lot more convenient then pulling
it/back it up/sending it somewhere, etc.

Spinrite is a tool no serious computer maintenance person should be
without. A bit pricey; but nothing else does the job like it does.

Won't *always* rescue a failing drive; but still ....

OTOH, once Spinrite *does* rescue my files, I still think it a good idea
to copy all of said files to a NEW drive, instead of continuing on with
one that's already failed once. The next time Spinrite might not be
able to rescue things. I've had that happen too.

Once a drive starts to go ....
And new drives are getting really cheap on a per-gigabyte cost-ratio.

 

[Bob Brown]

I decided to take the plunge and used spinrite, $80, and I found the tools,
like seatools and the ones from western digital are not very good. SR fixed
the problems my drive had and it was a lot more convenient then pulling
it/back it up/sending it somewhere, etc.

If mine is a physical problem then how would SR fix it?
What does SR exactly do for a drive?
thanks

 

[Michael Hawes]

I occasionally hear a CHIRP noise from my drive. It is more frequent
if I do a spyware scan or defrag, but it does happen when the comp is
idle like at night just before falling asleep.

SMART reports no errors or anything out of spec.
I've also downloaded several programs to "test" the drive and no
errors were reported.

I cannot run the WinXp GUI version of SEATOOLS but was able to burn a
cd dos version and run it and it also found zero problems/errors.

The drive is...
http://www.newegg.com/product/product.asp?item=N82E16822148140
Seagate Barracuda 7200.10 ST3320620AS (Perpendicular Recording
Technology) 320GB 7200 RPM 16MB Cache SATA 3.0Gb/s [I have it set to
SATA ONE because the motherboard does not support SATA II

Opinions/suggestions/conjecture welcome.
thanks.

Just backup your and keep an eye on it.

Mike.

 

[Frank McCoy]

If mine is a physical problem then how would SR fix it?
What does SR exactly do for a drive?
thanks

Spinrite goes out, reads and writes to a drive with internal
error-recovery disabled (as much as the interface allows) so it finds
errors; and then uses special reading and data techniques to recover
what would normally be unrecoverable data. This relies on knowing *how*
data is normally stored on a disk, the likeliest errors to occur with
that data, and similar methods.

It also, if you give it time, can read sectors, "scrub" them, and
rewrite the data cleanly. This helps as some data tends to get harder
to read over time, temperature, and drive aging.

All this (of course) takes a lot of time, usually many hours and
sometimes even days for the whole process to complete on a large drive.
However, you CAN pause, exit to reboot, restart, run Windows, and then
come back and continue where you left off even days later.

No other produce out there comes even close when it comes to recovering
damaged data, except for those *very* expensive companies that will tear
a drive apart and rebuild bad stuff inside.

However, once a disk starts crashing, even though Spinrite *might*
recover most of your data, it's time to get a new one before nothing
will work on the drive at all.

If you have spare time while not running programs though, Spinrite is
*wonderful* to run when you're gone, just as preventive maintenance.
The only problem is, it's stand-alone ... NO Windows or anything else
runs while it does. (That's needed for direct access to the drive
hardware.) You have to reboot to restart DOS, Unix, or Windows.

(Don't forget to hit "Escape" to let the program save where it's
currently working first though; rather than just hitting the RESET
button.)

 

[Ed Medlin]

If mine is a physical problem then how would SR fix it?
What does SR exactly do for a drive?
thanks

Just to add to what Frank M. said, SR will also isolate any bad sectors that
are not repairable and data will not be written to those areas anymore. HDDs
are finnicky animals and I have had SR repair drives I thought were almost
dead and they are still running and also ran it on drives that had very few
problems and failed within a week.......... You just never know. Once a
drive starts making strange noises, especially the well known "click of
death" is when I expect it to fail and get any valuable data off it. SR is
about the best drive analysis/maintainence software out there, but it can't
perform miracles........

Ed

 

[geoff]

That was my problem, my HD was not that old and I was getting SMART errors
on bootup. The WD tool said the drive has problems, the seatools said it
fixed the problem, but it did not.

I ran SR and it found a bad sector, marked it, and I never had a problem
since. However, if one is having serious problems, SR will show that in the
statistics window and recover data. The others are right, for serious
problems, switching out drives is best.

-g

 

[Franc Zabkar]

I decided to take the plunge and used spinrite, $80, and I found the tools,
like seatools and the ones from western digital are not very good. SR fixed
the problems my drive had and it was a lot more convenient then pulling
it/back it up/sending it somewhere, etc.

-g

I used to think that Spinrite was magic, but this piece by John Navas
gave me reason to think again:

http://grcsucks.com/spinrite.htm

For example, Steve Gibson claims that magnetic data decays over time
and needs to be refreshed. Navas counters this by saying that hard
drives rely on a dedicated servo surface, or embedded servo data, for
positioning purposes. This servo information is prerecorded at the
factory and cannot be rewritten in the field. If this data were
subject to decay, then hard drives would be failing in droves.

Having said that, I seem to recall that early MFM hard drives had a
"write precompensation" setting in the BIOS that told the controller
to allow for magnetic effects resulting from high data densities at
the inner cylinders. I believe this was to overcome migration problems
as tightly packed magnetic domains either repelled or attracted each
other.

In any case, no matter who you believe, I would think that
defragmenting your drive would probably refresh most of your disc's
surface.

- Franc Zabkar

 

[Frank McCoy]

In alt.comp.hardware.pc-homebuilt Franc Zabkar

I used to think that Spinrite was magic, but this piece by John Navas
gave me reason to think again:

http://grcsucks.com/spinrite.htm

For example, Steve Gibson claims that magnetic data decays over time
and needs to be refreshed. Navas counters this by saying that hard
drives rely on a dedicated servo surface, or embedded servo data, for
positioning purposes. This servo information is prerecorded at the
factory and cannot be rewritten in the field. If this data were
subject to decay, then hard drives would be failing in droves.

Having said that, I seem to recall that early MFM hard drives had a
"write precompensation" setting in the BIOS that told the controller
to allow for magnetic effects resulting from high data densities at
the inner cylinders. I believe this was to overcome migration problems
as tightly packed magnetic domains either repelled or attracted each
other.

In any case, no matter who you believe, I would think that
defragmenting your drive would probably refresh most of your disc's
surface.

It would, and does.
Spinrite's main advantage is in recovering data and finding bad spots to
mark unusable.

 

[Franc Zabkar]

In alt.comp.hardware.pc-homebuilt Franc Zabkar

It would, and does.
Spinrite's main advantage is in recovering data and finding bad spots to
mark unusable.

I confess that I'm a fan of Spinrite and Steve Gibson, and while I
actually dislike John Navas, I can see that he has some valid
questions. However, despite Navas' concerns, I have no doubt that
Spinrite can recover bad data by (1) repetitively rereading the bad
sector and hoping for one good read, or (2) by repetitively reading
the bad sector with ECC disabled and then reconstructing the data
using statistical methods.

As for Gibson's other methodology, let me start by saying that I was
first introduced to Spinrite in the early 90's when MFM stepper motor
drives were the norm. In those days you could actually low level
format a drive, set its interleave factor and write precompensation,
and you had to enter the bad block map by hand. I believe that many of
the claims that Gibson makes for Spinrite would have been applicable
in those days but maybe not today.

For example, Spinrite attempts to recover errant sectors by
approaching the track from both directions. The idea is that there is
some mechanical hysteresis which positions the R/W head slightly
off-track. However, AFAICS, modern drives have dual MR heads which
don't really have this problem because the read head is narrower than
the write head. See Fig 1 in the following article:

http://baden.nu/public_html_index/S...MR_Head/Magnetoresistive_Head_Technology.html

Secondly, there is Gibson's claim that Spinrite can fool the drive's
AGC circuitry into lowering the gain of the read amplifier by
preconditioning it with a best case data pattern, ie one that produces
the strongest possible read amplitude. I'm no expert, but AFAICS, the
AGC level would be largely determined by the prerecorded embedded
servo, gap, ID, and sync bytes immediately preceding each data field.

See http://www.hitachigst.com/hdd/ipl/oem/tech/noid.htm

Then there is the claim by Navas that "Steve encouraged people to use
SpinRite to 'recover' areas that had been detected and marked as
defective at the factory". Clearly this is not possible with today's
drives because sector sparing is carried out transparently by the
drive's microcontroller, but even if it were possible it would be a
stupid thing to do. In fact one of Spinrite's reasons for existence is
to locate marginal sectors and take them out of service.

- Franc Zabkar