Steve Smith <(E-Mail Removed)> wrote in
message news:OLGdb.28636$0Z5.19030@lakeread03...
> I have a Dell Inspiron 8100 laptop with an IBM Travelstar
> hard drive running Windows XP Home that suddenly started
> working EXTREMELY SLOWLY. It was like running XP on
> an old 386 with 16MB memory. The computer has a 1Ghz CPU
> and 256MB memory. It took 10 to 15 minutes to open MS Word.
> Sometimes it takes 5 minutes for the menu to appear after I click
> on the Start button. It'd take it almost 15 minutes to boot!
The usual cause of that effect, particularly when replacing the
hard drive fixes that, is that the bad one is having problems
reading the data off the platters, but does eventually succeed
on retrys. That retrying slows things down very dramatically.
> I tried every fix imaginable and I finally determined that the
> hard drive is bad. I made this determination by "Ghosting" to
> a different hard drive and then replacing the suspect drive with
> the Ghosted drive. It runs great with the new hard drive. For
> whatever reason the old hard drive passed Dell's diagnostic
> program for this laptop, PC Certify diagnostic software
Yeah, those often dont notice retrys that succeed. Neither does scandisk.
> and the IBM's "Fitness Test" hard drive diagnostics.
Thats pretty sad.
Have you tried one of the SMART tools to
see if the drive has recorded lots of retrys ?
> I even did a low level format and a clean
> install of Windows but that didn't help.
How did the install actually go ? Was it noticeably slow to install ?
> A replacement hard drive worked like a charm. I guess
> you can't always trust hard drive diagnostic software.
Yep, no substitute for swapping hardware when it comes to the crunch.
> Perhaps it's possible for a hard drive to slow down because
> of mechanical reasons maybe due to bad bearings
It wont be that. It wont be rotating slowly.
> or something
Yep, most likely continual retrys which do succeed.
> and maybe this doesn't get tested by drive diag software
Yes, most of it doesnt test for retrys that succeed.
It would be interesting to see what the HDTach results are.
> or "SMART" analysis. Has anyone seen this problem before?
Yes, its something that the industry has been stuck with forever.
SMART should at least in theory have picked up
the problem. Have you actually got that enabled ?
|