View Single Post
  #32  
Old October 29th 17, 04:51 AM posted to alt.comp.os.windows-8
Paul[_32_]
external usenet poster
 
Posts: 11,873
Default Asus X550J laptop

Mayayana wrote:
"Paul" wrote

| Now, what's wrong with that "theory". Well, on the Seagate
| drives I've got, I've *never* seen Current Pending Sector
| go non-zero. Even when other activity indicates the drive
| is sick, and Current Pending should be growing. Some brand
| of drive, probably is using Current Pending Sector, but
| not in the case of the Seagates I've owned.
|
| Current Pending returns to zero, if an opportunity comes
| along to write the entire drive.
|
| Reallocated Sector Count is a measure of how many spares
| have been used up. It's thresholded, so only after a large
| number of sectors were spared, does the count value go non-zero.
| The result is, the user is unaware exactly how large the
| spared sector count is.

And that's with all drives? All SMART drives? I'm
not clear about the context here. It sounds like
you're saying that with recent vintage drives the
health reports can't be trusted. That isn't really news,
is it? Does that have any connection to the
OS/Windows version?


All IDE/SATA drives now have SMART.

It provides statistics.

The drive tests itself occasionally, though I've never had
a drive wobble enough to fail on a self test and trigger
the BIOS-level warning. The BIOS on quite a few motherboards,
reads SMART at startup, and is supposed to be able to stop the
boot and warn you that the hard drive is sick. You don't absolutely
need to keep HDTune loaded, if you have a BIOS setting to warn
you that a drive is unhealthy. Naturally, the health calculation
leaves something to be desired (see picture below for why that
can be - not everyone agrees on how to interpret that screen).

The behavior of Reallocated Sector Count is necessary because
of human nature.

If the Reallocated statistic reported the actual count, people
buying new hard drives would be ****ed, that the counter is
always non-zero from the factory. The platters always have
tiny defects. And sectors are spared out when the drive leaves
the factory.

The factory has an "acceptance" criterion. Say the acceptance
is 100,000 sectors spared before it leaves the factory. Then
the statistic will read 0 until the actual number of spared
out sectors surpasses 100,000. Then, the "lifetime" percentage
in that field, will drop from 100% life to 0% life, as the
number of reallocations changes from 100,000 to 105,500.

That's what I mean by thresholded. They don't want you cherry
picking drives, and sending them back to Newegg if the statistic
is 2000 from the factory. If it's only 2000, then it leaves
the factory reading zero. If 98,000 more happen while you're
using it, it will finally have a non-zero count. It has a
further capacity of around maybe 5500 or so. In a relatively
short time, you could see the percentage value dropping and
realize "hey, I'd better do something".

This is all supposition on my part, as the manufacturer is not
going to admit to this. It doesn't take too many grain defects
in the platters of a 4TB drive, to create a need to spare out
a sector. And they don't sit there tossing platters into
a huge pile in the corner, because the drive won't read zero.
They allow the drive to have a certain number of reallocations
before it leaves the factory. Drives were leaving the factory
with 100,000 defects in the 9GB drive era.

I've had some drives, that were flaky enough, that they needed
to be "written from end to end" at least one, just to make
their transfer rate performance consistent. All the drives I've
bought in the last four years or so, don't have that behavior.
The only egregious stuff now, is FDB motors making funny sounds
at shutdown (lubrication starvation on the shaft). And the excessive
spindowns that even the expensive drives are doing (that I'm not
able to turn off!). I hate that spindown crap, and that's why I
was buying more expensive drives - not because they last longer,
but because of their "less-compromising" behaviors.

This is an example of the way drives should be built. It has
37000+ hours of power-on life on it. It spins constantly and
*never* spins down or parks the heads. It's always ready when
I search against it. It shows zero for the two indicators I
use for health. I will not get another like this, for as long
as I live. It just keeps going and going. It holds a copy of Win2K,
to give some idea how crusty the content is :-) Keeping it
spinning is a Smithsonian experiment of mine... I'm really
curious how long it can continue like this. And yes, I do occasionally
backup and restore, just to make sure the drive isn't "cheating"
in any way (as a check for latent faults).

https://s1.postimg.org/4l2b9u5eb3/golden_HDD.gif

Paul
Ads