Ok, I am a post to the list because I think that's no problem.

On Thu, 19 Apr 2007 23:24:23 +0400
Aaron Kulkis <[EMAIL PROTECTED]> wrote:

> eshsf wrote:
> > Hello,
> > 
> > On Tue, 17 Apr 2007 22:27:36 -0500
> > "Rajko M." <[EMAIL PROTECTED]> wrote:
> > 
> >> On Tuesday 17 April 2007 04:15, Clayton wrote:
> >> ...
> >>> I'm assuming the drive is unrecoverable (ie it's not worth risking my
> >>> data to try and run the Maxtor tools on it and "repair" the drive)...
> >>> so it's up for replacement in the next week or so.
> >>>
> >> You may try to slow down the drive. 
> >> I got one that was about to fail, but decreasing the speed, using Maxtor 
> >> utilities, from 66 to 33 helped and it still runs. 
> > 
> > Probably, the decreasing speed problem is because the HDD controller
> > changed a bad sector into a spare sector. and the spare sector possibility
> > don't optimize to seek, I guess.
> > 
> > AFAIK, A modern HDD has a number of spare sectors. and when happened a bad
> > sector, that can replace a bad sector into a spare sector by using the hard
> > drive tools, etc.
> > *But*, there is a limitation in the number of spare sectors (and depend on 
> > HDD).
> > 
> 
> Assuming a 15ms seek time, I sincerely doubt that the original poster
> would have noticed the effects of this... that would account for a
> whopping 30ms (i.e. 0.03 seconds)...and once the data is read
> into memory, it's done ... that disk block is now in the I/O
> buffers, and there is no need to revisit it.

exactly. when I did post, I didn't consider it at all, almost all the HDD
have the I/O buffers on drive and the Linux have a buffer cache though.

Yes, so, only when the first accessing it or flushed a cache, it might influence
but the influence is a little bit.



Thanks,
eshsf

-- 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to