Re: Journaling pointless with today's hard disks?

[Jeff V. Merkey]

Check out the hotfixing code in NWFS.  It handles exactly what this 
long and drawn out thread has discussed, and it's already in 
Linux.  The code is contained in nwvp.c.  I can tell you 
that in the past three years of running NWFS on Linux and all
the time I worked at Novell from about 1996 on, I never once saw 
a server hotfix data after the newer "data guard" drive technologies
came out.  In fact, by default, I make the hotfix area on the drive
about .1 % of the total space, since it;s probably just wasted 
space at this point.  

It's just wasted space these days, but it is a good idea to keep it 
around, just in case the "pointless" argument turns out not to 
be pointless and someone gets eaten by a shark (1 in 100,000,000) at
the same instant they are struck by lightening (1 in 200,000,000).

:-)
Jeff

On Thu, Nov 29, 2001 at 11:21:57PM +0100, Pavel Machek wrote:
> Hi!
> 
> > > Assuming the drive's inherent bad-block detection mechanisms don't find it 
> > > and remap it on a read first, rapidly consuming the spare block reserve.  But 
> > > that's a firmware problem...
> > 
> > Drives should never reassign blocks on read operations, because they'd
> > take away the chance to try to read that block for say four hours.
> 
> Why not? If drive gets ECC-correctable read error, it seems to me like
> good time to reassign.
> 								Pavel
> -- 
> "I do not steal MS software. It is not worth it."
>                                 -- Pavel Kankovsky
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [e-mail]
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/