Re: PC grinding to a halt - help!

The message <nbudnQfm9cNFtQXanZ2dnUVZ8vSdnZ2d@xxxxxx>
from "Ginchy" <jow@xxxxxxxxxxxxxxxxx> contains these words:


====huge snip====

Wait a minute.....is it the sata cable? On my pc I sometimes get a situ
where the cable gets nudged slightly and although its not actually out of
the socket its not right. What happens then is that as you access the drive
it sometimes takes ages to find it and sometimes it would vanish
altogether.
Nudging the cable would do the trick. This could account for the
slowness as
your system tries to find the drive. Try pushing the sata cables in both at
the mb end and the hd end and see if theres a difference.

If its not that at least you now have the case open and can remove that
modem :o)

Look for an excellent utility called hdtune which will do a thorough
diagnosis of your drive/s including read/write performance.

Ever since I changed over to sata hdds with a new build nearly 4 years
ago, I've noticed inexplicable hangs every so often (which, eventually,
proved to all involve hdd data transfer activity).

Once I had eliminated memory and cpu issues as possible causes, and had
widened my horizons to include the flakiness of the sata connector as a
potential source for such hangs, I found that on subsequent hang events,
I could 'unhang' the PC by the expedient of giving the sides of the PC
box a "Jolly Good Thump" or two, thus proving that a bad contact was
involved.

The sata connectors were the prime suspect in this scenario for two
reasons. The first, very obviously, was on account of the very poor
mating of the connectors, particularly at the hdd end, exposing the
connection to a much higher risk of intermittent contact problems. The
second was the fact that this link (unlike the pata one) employed error
detection/correction which would prevent command and data errors from
corrupting file transfers and (most importantly) total corruption of the
existing data structure held on the disk drive.

IOW, such data cable problems could only cause nothing more harmful
than hangs or delays (the very symptoms I was witnessing).

Such 'bad contact' problems in the older pata setup could produce much
more 'tragic' consequences, but, the generally better connector
reliabilty made such tragedies a rarity. The fact that sata employs such
error detection is the only saving grace that has allowed such a weak
connector to be used at all.

The main part of the weakness of the sata connector is largely down to
its envisaged use in 'hot swap' bays where the sata plug is not on the
end of a 'fly lead' but part of the carrier where the 'wobbly' plug
syndrome simply doesn't exist.

I currently have a customer's Presario box on my bench which was
failing to boot into windows due entirely to FS corruption (a chkdsk /r
resolving this issue). Whilst clearing out the numerous spybot/adware
parasites from each of the 3 user login accounts, the system 'hung' with
the maxtor sata drive making terminal sounding clicking noises. In this
case a "Jolly Good Thump" or two didn't help but I was able to shut the
PC down and reseat the sata cable at the hdd end and resume without any
signs of FS damage.

However, a little later on, the symptoms returned. This time I decided
to swap the sata cable ends over. Unplugging from the MoBo header proved
difficult on account the header socket incorporated the locking
retaining feature which required a screwdriver to depress the release
tab (elevate, in reality, due to the MoBo idiot designer not considering
the base of a tower case becoming an obstacle).

I decided to remove the cable tie the idiot system builder had used to
'tidy up' the slack in the sata cable in order to reduce the 'pull out
forces' this had created on the hdd connector. This seems to have cured
the "terminal sounding clicking noise" issue with the maxtor.

I haven't run any hard disk diagnostics simply because, the alarming
sounding disk noises, if associated with terminal impending failure,
would not have permitted trouble free booting back into windows, nor
allowed either a scan for FS errors to complete or the defrag utility to
succeed.

However, I can't totally preclude there being a hdd internal component
failure brewing up (it is a Maxtor drive after all!). Which is why I've
scheduled a full disk scan on next boot... which has hung at 4% test 4
of 5 :-(

I'm now running memtest (a very unlikely culprit, but best 'Leave no
stone unturned') which I'll follow up with a hdd diagnostic scan (the
noises sound like the drive is parking the heads as for a reset and, it
now seems extremely unlikely to be a sata connector issue - if it ever
was).

Memtest passed without error (as it must have done when I almost
certainly ran this check on receipt of the system box last week). I'm
now running the seagate/maxtor utility which has (as per usual with the
seagate utility) reported the drive as having been 'over temperature 253
deg! Yeah, right! However, it _has_ failed the short test.

In view of the lack of the more usually obvious symptoms such disk
drive faults tend to generate, a surprising result. However, a definite
(and consistent) fail result does at least remove the guesswork from the
equation.

--
Regards, John.

Please remove the "ohggcyht" before replying.
The address has been munged to reject Spam-bots.

.