On Fri, 27 Aug 2004 21:11:16 -0400, "Ritter197"
How can I find out whether there are bad clusters on a harddrive?
Best is to test the HD with an OS-agnostic diagnostic. You cannot
trust the OS with this stuff, especially if NTFS.
If not NTFS, you can use DOS mode Scandisk and surface scan. That
shows a map of cluster space with previously "fixed" bad clusters
visible - plus you can see latency when the HD's firmware tries to
"fix" bad sectors (thus hiding impending failure from you).
But Scandisk can't test NTFS, and ChkDsk /R is even more aggressive in
hiding impending failure from you. No prompts on errors, it "fixes"
automatically - plus, you have two levels of on-the-fly failure-hiding
systems, the HD's own firmware plus NTFS's on-the-fly "fixing".
That being the case, NTFS is more likely to have hidden bad sectors
from you, and so it becomes more important to see what has been going
on - by reading SMART's logs of what has been "fixed" at the HD
firmware level at least. The HD vendor's web site usually has free
tools that gives a "quick" test (just an editorialized summary of
stored SMART logs) plus a longer test that actually tries reading the
HD disk surface, rather than looking only at SMART logs.
OTOH, some 3rd-party utils such as AIDA32 give you more detailed SMART
info; e.g. instead of "normal", an actual error count.
Yet the service man told my daughter, her harddrive had bad clusters which
every now and then gave her a blue screen and said the system has to shut
down.
Bad clusters are like small strokes. You will lose data,
functionality, possibly the whole installation, if you let them bite.
I have that HD now in my external housing and see nothing bad, but I might
be missing something.
Those are highly at risk, because:
- wobbly power and cabling
- poor ventillation
- risk of being dropped
If you want to kill a HD, then cook it, spike the power to it, and
spin it up and down a lot.
--------------- ----- ---- --- -- - - -
Who is General Failure and
why is he reading my disk?