> How about a simple utility, added to the ext2 utilities, that compacts
> directories that can be run periodically as the user wishes. By the way
> don't these ghost entries hamper performance, albeit possibly by a
> negligible amount. I would think that these ghost entries get processed
> along with the other entries whenever the directory is accessed. And if
> this utility is made other file system optimization options can be thrown
> in. Perhaps the directory entries can be sorted by their last access date
> as the directory is compacted so as to improve file access performance.
>
> -----Original Message-----
> From: Ted Deppner <ted@jasmine.psyber.com>
> To: Chad C Giffin <typo@t-net.org>; linux-kernel@vger.rutgers.edu
> <linux-kernel@vger.rutgers.edu>
> Date: Thursday, July 16, 1998 1:27 PM
> Subject: Re: ext2fs enhacement/fix [re: shrinking directories]
>
>
> >On Thu, Jul 16, 1998 at 12:33:50AM -0400, Chad C Giffin wrote:
> >> Although it is really not a problem for most that directory files do not
> >> shrink on removal of directory entries, it still is a problem for some
> and
> >> should not be ignored IMO.
> >>
> >> I'd like to propose a solution that is not too complicated.
> >
> >[acceptable solution (if properly implimented) deleted]
> >
> >I'd like to pose the question, Is this really necessary?
> >
> >On your home machine I can't imagine more than 1mb on the average system
> being
> >wasted to this excess.
> >
> >On a server, I can't imagine _wanting_ to recover space (under any scheme)
> >when most servers have a cyclic nature, and even if space were saved for a
> >period of time, there was a time (hours or days ago) when the directory
> NEEDED
> >to be that large.
> >
> >And on busy servers, where you might have 20-30 mb tied up this way (a news
> >machine just after expires), within 12 to 24 hours they'll be full again.
> >
> >In any case, I always have 100's of extra MBs of drive space laying around.
> >Even if I could recover 50mb for 12 hours or a week, it I wouldn't care,
> it's
> >such a minor thing.
> >
> >--
> >Ted Deppner
> >ted@psyber.com
> >http://www.psyber.com
> >
> >-
>
--
marty
leisner@sdsp.mc.xerox.com
The Feynman problem solving Algorithm
1) Write down the problem
2) Think real hard
3) Write down the answer
Murray Gell-mann in the NY Times
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html