Re: defrag and error checking

From: vampyre wolf <vampyrewolf_at_no.spam.please>
Date: Thu Aug 23 2007 - 18:13:27 CST

The list of glitches:
1. If the system isn't left for a minute after logging in, and beryl is
started, my KDE desktop crashes and loads gnome (without forcing me to log
in again).
2. Running kaffiene, kbounce and firefox (typical load when I get home from
work)... quite often the screens for kbounce show up blank, or a menu on
kaffiene shows up blank.
3. Using alt-F2 to load programs, probably 75-80% of the time I don't get
any text showing up (although the drop down menu of previous entries works).

Upgraded my HD/RAM when compu-smart shut down. The problems have only been
showing up the last week or so which is why I thought error checking &
defrag. Didn't think much the first couple times it showed up because I was
also burning a few cd's and had a decent load on my computer. But when the
glitch shows up daily for a week, there's something wrong.

On 8/23/07, Dylan Griffiths <dylang@thock.com> wrote:
>
> Kurtis Peterson wrote:
> > Error checking comes in the form of fsck but there is no defrag as linux
> > (and more specifically the ext 2/3 file system) is a "neater" way to
> > write files to the drive and theoretically doesn't need to be
> > defragged. Windows and the Fat or NTFS file systems can be very messy.
> > They start writing new files in the first availible free space
>
> This is not (strictly) true.
>
> All file systems attempt to allocate contiguous blocks. Due to the
> nature of how FAT works, lack of contiguous blocks = lots of HD syncs
> since it's just a linked-list of blocks.
>
> NTFS and EXT[23] use trees which allow for minimal seeking. There is
> little/minimal overhead from breaking up files (if need be). NTFS and
> EXT[23] should not require defragmentation if you keep enough free
> space. As files are read/written, the FS logic will reallocate space to
> reduce fragments as possible. In this regard, NTFS is much better than
> FAT.
>
> > regardless of whether that free space is big enough to handle the file
> > or not. After time this requires you to Defragment the drive by moving
> > files around so they are contiguous on the drive. Linux doesn't work
> > this way and has background daemons which are constantly optimizing the
> > drive with spare CPU cycles so you really don't need to. Give fsck a
>
> You have to run e2fsdefrag or the like if you want that -- it's not a
> daemon, but part of the FS layer of the kernel.
>
> > try and if that doesn't work it probably means there is another issue
> > and it's not your file system.
>
> I would suspect hardware issues. Ubuntu isn't Slackware-stable, but I'd
> still go with my gut of marginal RAM/CPU over a software issue on the
> Linux side if you are experiencing true random/unexplained behaviour.
> If you're experiencing easy-to-repeat behaviour, that's a software bug
> somewhere. Windows drivers are usually so poorly written that this
> leads to lots of troubles (ATI => 0x000 STOP stuff, etc).
>
> Defragmentation will have no effect on a computer in terms of stability,
> only in terms of IO throughput, and then only on poorly designed (FAT 16
> or 32) file systems.
>
> --
> I'm interested in upgrading my 28.8 kilobaud internet connection to a
> 1.5 megabit fiberoptic T1 line. Will you be able to provide an IP
> router that's compatible with my token ring ethernet LAN configuration?
>
> To unsubscribe, send a message with the word "unsubscribe" (without the
> quotes) in the body to linux-request@slg.org
> Archives are at http://list.slg.org/
>
Received on Thu Aug 23 18:13:40 2007

This archive was generated by hypermail 2.1.8 : Thu Aug 23 2007 - 18:13:43 CST