Low level formatting - [was Re: slow (s-l-o-w) install (TRY)]

Bill Davidsen davidsen at tmr.com
Mon Feb 18 17:54:48 UTC 2008

Mikkel L. Ellertson wrote:
> Alan Cox wrote:
>> On Sat, 16 Feb 2008 19:54:52 -0600
>> Frank Cox <theatre at sasktel.net> wrote:
>>> Low level what?  If you mean low-level formatting, I don't think 
>>> that's been
>>> possible to do with hard drives since the days of RLL and MFM.
>> Correct (well the very earliest days of IDE), and it wouldn't affect
>> performance anyway.
> I can remember people killing the early IDE drives by low-level 
> formatting. The tools for low-level formatting MFM/RLL drives would mess 
> up the IDE drives. With later drives, they would ignore the low-level 
> formatting attempts.
> Themain reason for low-level formatting MFM/RLL drives was moving them 
> to a new controller, or changing the interleave - nether are applicable 
> to modern IDE/ATA drives. (It was easier to low-level format then to 
> tweak the controller data clock to match the old controller.)
The main reason indeed, by going to a different interleave matched to 
your usage you could get a big jump in performance (or make it dog 
slow). And by playing with large sector sizes you could get a large jump 
in capacity. Just as a 3-1/2 floppy can jump from 1440k to 1920k with 
large sectors, some hard drives could get a similar boost in capacity, 
up to 40%.

I wouldn't try that today with zoned formatting, more CPU in the on-disk 
controller, etc, etc. Actually, I would send the LL format command to 
the drive if I know what it was, if the controller does it, it would 
probably work as well as it did at the factory.

Bill Davidsen <davidsen at tmr.com>
   "We have more to fear from the bungling of the incompetent than from
the machinations of the wicked."  - from Slashdot

More information about the fedora-list mailing list