Mayayana wrote:
"philo" wrote

| I was able to test RAM. That checked out. I ended
| up installing it into a Win7 box and running Hiren's
| boot disk. The WD diagnostic came out with error 7
| and quit. BootIt sees all the partitions, but the data
| on them seems to be limited. Chckdsk retrieved all sorts
| of things on the Windows partition but couldn't access
| any of the others.
| At this point I'm thinking there must be a problem
| with the hard disk.
|
| Yes, indeed it does sound like a HD issue

On 2 separate runs from a boot disk the WD utility
came up with error 7 and then while doing a "media
scan", in preparation for a thorough check, it stopped
with error 225. Their error page says that means,
"Too many errors to continue" and advises getting
a new disk.

Meanwhile various tools said that it passed a
SMART scan.

This whole thing has got me to check out SMART for
perhaps the third time, and I've come away with the
same impressions I've had in the past: It seems to be
of little value and I've yet to find a clear explanation
of how to interpret it. People recommend a Wikipedia
page, but that's not very helpful. Even the categories
reported vary between tools.

" the research showed that a large proportion (56%)
of the failed drives failed without recording any count
in the "four strong S.M.A.R.T. warnings" identified as
scan errors"
"36% of drives failed without recording any S.M.A.R.T.
error at all, except the temperature, meaning that
S.M.A.R.T. data alone was of limited usefulness in
anticipating failures."

I'm still very curious about how the disk could die in
3 years, but statistically that's not unheard of. And
I don't know how it was used. I guess the only thing
I can do is to reinstall and try to minimize unnecessary
background junk when I do the setup. That's the one
aspect that's got me suspicious. When I search for
links about hard disk trouble I seem to find a lot of
complaints from people about ceaseless activity, which
they eventually trace to some unnecessary 3rd-party
applet.

The "reallocated" metric works best for error patterns
spread uniformly over the disk surface.

However, that's not the only failure pattern. I had
a disk here, with an obvious "slow patch" which means
read errors and re-allocated galore. And the thresholded
reallocated data field still said "zero", implying 100%
health. If I had been using automated surveillance,
it would have missed the warning signs. However, all
the OS files were slow to load, so a human could sure
tell something was wrong. (That was for a 60GB OS
partition, slow as molasses, on a 500GB drive with
440GB of "good" sectors. An HDTune benchmark showed
the problem for what it was. A wide bad spot.)

I use SMART, in combination with common sense. If
I see, hear, or smell trouble, I get out the SMART
panel and have a look, for confirmation. I won't
always get an "indicator" from SMART, but it's better
than nothing. And if something is registering,
I can take note of the degradation rate. The
growth rate of the reallocated, on a day-by-day
basis. That tells me how much trouble I'm in,
and how fast I should run to the store for
a spare drive.

SSDs are slightly different, and because of the evil
end-of-life policy of some brands, maintenance
should be taken more seriously. (An SSD drive can simply
stop responding to your queries, read or write,
as a "service" to you!) A HDD won't do that.
When you buy an SSD, *always* check the web for
info on end-of-life behavior.

Paul